CN106227834B - Multimedia resource recommendation method and device - Google Patents

Multimedia resource recommendation method and device Download PDF

Info

Publication number
CN106227834B
CN106227834B CN201610594958.7A CN201610594958A CN106227834B CN 106227834 B CN106227834 B CN 106227834B CN 201610594958 A CN201610594958 A CN 201610594958A CN 106227834 B CN106227834 B CN 106227834B
Authority
CN
China
Prior art keywords
multimedia
class
multimedia resource
recommended
resources
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.)
Expired - Fee Related
Application number
CN201610594958.7A
Other languages
Chinese (zh)
Other versions
CN106227834A (en
Inventor
滕飞
赵磊
单明辉
尹玉宗
姚键
潘柏宇
王冀
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Alibaba China Co Ltd
Youku Network Technology Beijing Co Ltd
Original Assignee
Youku Network Technology Beijing Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Youku Network Technology Beijing Co Ltd filed Critical Youku Network Technology Beijing Co Ltd
Priority to CN201610594958.7A priority Critical patent/CN106227834B/en
Publication of CN106227834A publication Critical patent/CN106227834A/en
Application granted granted Critical
Publication of CN106227834B publication Critical patent/CN106227834B/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

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

Abstract

The invention relates to a method and a device for recommending multimedia resources. The method comprises the following steps: dividing multimedia resources to be recommended into first-class multimedia resources and second-class multimedia resources; determining a click rate pre-estimated value of the first type of multimedia resources according to historical click data of the first type of multimedia resources; searching a first class of multimedia resources matched with a second class of multimedia resources, calculating the matching degree of the second class of multimedia resources and the matched first class of multimedia resources, and calculating the click rate estimated value of the second class of multimedia resources according to the matching degree and the click rate estimated value of the matched first class of multimedia resources; and recommending the multimedia resources to be recommended according to the click rate estimated value of the multimedia resources to be recommended. According to the method and the device for recommending the multimedia resources, the newly uploaded multimedia resources can be also recommended at an opportunity, so that the recommendation accuracy of the multimedia resources is improved, and the recommendation effect of the multimedia resources is improved.

Description

Multimedia resource recommendation method and device
Technical Field
The invention relates to the technical field of information, in particular to a method and a device for recommending multimedia resources.
Background
The existing video recommendation technologies mainly include a click rate-based video recommendation technology, an uploading time-based video recommendation technology and a text information-based video recommendation technology. Wherein, the text information refers to the title of the video and/or the description information added by the uploader for the video. In the existing click-through-volume-based video recommendation technology, some newly uploaded videos with small click-through volume are difficult to recommend. The existing video recommendation technology based on the uploading time does not consider the popularity of the video, and the recommendation effect is poor. In the existing video recommendation technology based on text information, the accuracy of video recommendation is low due to the fact that the relevance of the text information of some videos and video contents is low.
Disclosure of Invention
Technical problem
In view of the above, the technical problem to be solved by the present invention is to solve the problem that the existing multimedia resource recommendation technology has a poor recommendation effect.
Solution scheme
In order to solve the above technical problem, according to an embodiment of the present invention, a method for recommending multimedia resources is provided, including:
dividing multimedia resources to be recommended into first-class multimedia resources and second-class multimedia resources;
for each first-class multimedia resource, determining a click rate pre-estimated value of the first-class multimedia resource according to historical click data of the first-class multimedia resource;
for each second-class multimedia resource, searching a first-class multimedia resource matched with the second-class multimedia resource, calculating the matching degree of the second-class multimedia resource and the matched first-class multimedia resource, and calculating the click rate estimated value of the second-class multimedia resource according to the matching degree and the click rate estimated value of the matched first-class multimedia resource;
and recommending the multimedia resource to be recommended according to the click rate estimated value of the multimedia resource to be recommended.
For the above method, in a possible implementation manner, dividing the multimedia resources to be recommended into a first class of multimedia resources and a second class of multimedia resources includes:
respectively acquiring historical click data of each multimedia resource to be recommended;
and for each multimedia resource to be recommended, judging whether the current click rate of the multimedia resource to be recommended is in a decline period according to the historical click data of the multimedia resource to be recommended, if so, determining the multimedia resource to be recommended as a first class of multimedia resource, and otherwise, determining the multimedia resource to be recommended as a second class of multimedia resource.
For the above method, in a possible implementation manner, dividing the multimedia resources to be recommended into a first class of multimedia resources and a second class of multimedia resources includes:
respectively acquiring the uploading time of each multimedia resource to be recommended;
and for each multimedia resource to be recommended, judging whether the time length of the uploading time of the multimedia resource to be recommended from the current system time is greater than a first preset value, if so, determining the multimedia resource to be recommended as a first class of multimedia resource, and otherwise, determining the multimedia resource to be recommended as a second class of multimedia resource.
For the above method, in a possible implementation manner, for each first-class multimedia resource, determining a click rate pre-estimated value of the first-class multimedia resource according to historical click data of the first-class multimedia resource, includes:
for each first-class multimedia resource, acquiring historical click data of the first-class multimedia resource;
training the time attenuation coefficient of the first type of multimedia resources according to the historical click data of the first type of multimedia resources;
and calculating the click rate pre-estimated value of the first type of multimedia resources according to the historical click data of the first type of multimedia resources and the time attenuation coefficient.
For the above method, in a possible implementation manner, training a time attenuation coefficient of the first type multimedia resource according to the historical click data of the first type multimedia resource includes:
training a time attenuation coefficient theta of a first-class multimedia resource i by adopting a formula 1 according to historical click data of the first-class multimedia resource i, wherein the historical click data of the first-class multimedia resource i comprises the click quantity of the first-class multimedia resource i before a specified date;
f(t)＝Vi0×eθtformula 1;
wherein, Vi0Representing the click rate of the first type multimedia resources i on the specified date; t (0) represents the specified date, and f (0) represents Vi0(ii) a When t > 0, t represents the tth day before the specified date, f (t) represents the click rate of the first-class multimedia resources i on the tth day before the specified date; when t is less than 0, t represents the day-tth after the specified day, f (t) represents the click rate estimated value of the first-class multimedia resources i on the day-tth after the specified day;
calculating the click rate pre-estimated value of the first type of multimedia resources according to the historical click data of the first type of multimedia resources and the time attenuation coefficient, wherein the method comprises the following steps:
and calculating the click rate estimated value of the first-class multimedia resource i on the-t day after the specified date by adopting the trained formula 1.
For the above method, in a possible implementation manner, training a time attenuation coefficient of the first type multimedia resource according to the historical click data of the first type multimedia resource includes:
training a time attenuation coefficient theta of a first-class multimedia resource i by adopting a formula 2 according to historical click data of the first-class multimedia resource i, wherein the historical click data of the first-class multimedia resource i comprises the click quantity of the first-class multimedia resource i before a specified date;
g(t)＝lg(Vi0×eθt) Formula 2;
wherein, Vi0Representing the click rate of the first type multimedia resources i on the specified date; t-0 indicates the specified date, and g (0) -lgVi0(ii) a When t > 0, t represents the t day before the specified date, g (t) represents the logarithmic value of the click rate of the first-class multimedia resources i on the t day before the specified date; when t is less than 0, t represents the day-tth after the specified date, g (t) represents the logarithm value of the click rate estimated value of the first-class multimedia resources i on the day-tth after the specified date;
calculating the click rate pre-estimated value of the first type of multimedia resources according to the historical click data of the first type of multimedia resources and the time attenuation coefficient, wherein the method comprises the following steps:
and calculating the logarithm value of the click rate estimated value of the first-class multimedia resource i on the-t day after the specified date by adopting the trained formula 2.
For the above method, in a possible implementation manner, for each second-class multimedia resource, searching for a first-class multimedia resource that matches the second-class multimedia resource, and calculating a matching degree between the second-class multimedia resource and the matched first-class multimedia resource, includes:
for each second-class multimedia resource, searching a first-class multimedia resource matched with the second-class multimedia resource according to the specified information of the second-class multimedia resource; the specifying information includes at least one of: uploading time, time length and uploader information;
and calculating the matching degree of the second type of multimedia resources and the matched first type of multimedia resources according to the specified information of the second type of multimedia resources and the specified information of the matched first type of multimedia resources.
For the above method, in a possible implementation manner, the click rate estimated value of the second type of multimedia resource is calculated according to the matching degree and the click rate estimated value of the matched first type of multimedia resource, specifically:
and taking the product of the matching degree and the click rate estimated value of the matched first-class multimedia resource as the click rate estimated value of the second-class multimedia resource.
For the method, in a possible implementation manner, recommending the multimedia resource to be recommended according to the click rate estimate value of the multimedia resource to be recommended includes:
sequencing all the multimedia resources to be recommended according to the sequence of the click rate estimated values of the multimedia resources to be recommended from high to low;
and taking out M multimedia resources to be recommended which are sequenced at the front from all the sequenced multimedia resources to be recommended for recommendation, wherein the size of M is preset.
In order to solve the above technical problem, according to another embodiment of the present invention, there is provided a recommendation apparatus for multimedia resources, including:
the system comprises a dividing module, a recommending module and a recommending module, wherein the dividing module is used for dividing the multimedia resources to be recommended into a first type of multimedia resources and a second type of multimedia resources;
the first click rate pre-estimated value determining module is used for determining the click rate pre-estimated value of each first type of multimedia resource according to historical click data of the first type of multimedia resource;
the second click rate pre-estimated value determining module is used for searching a first class of multimedia resources matched with the second class of multimedia resources for each second class of multimedia resources, calculating the matching degree of the second class of multimedia resources and the matched first class of multimedia resources, and calculating the click rate pre-estimated value of the second class of multimedia resources according to the matching degree and the click rate pre-estimated value of the matched first class of multimedia resources;
and the recommending module is used for recommending the multimedia resource to be recommended according to the click rate estimated value of the multimedia resource to be recommended.
For the apparatus, in a possible implementation manner, the dividing module includes:
the historical click data acquisition submodule is used for respectively acquiring the historical click data of each multimedia resource to be recommended;
and the first dividing module is used for judging whether the current click rate of the multimedia resources to be recommended is in a decline period or not according to the historical click data of the multimedia resources to be recommended for each multimedia resource to be recommended, if so, determining the multimedia resources to be recommended as first-class multimedia resources, and otherwise, determining the multimedia resources to be recommended as second-class multimedia resources.
For the apparatus, in a possible implementation manner, the dividing module includes:
the uploading time obtaining submodule is used for respectively obtaining the uploading time of each multimedia resource to be recommended;
and the second division submodule is used for judging whether the time length of the uploading time of the multimedia resource to be recommended from the current system time is greater than a first preset value or not for each multimedia resource to be recommended, if so, determining the multimedia resource to be recommended as a first class of multimedia resource, and otherwise, determining the multimedia resource to be recommended as a second class of multimedia resource.
For the above apparatus, in one possible implementation manner, the first estimated click quantity value determining module includes:
the historical click data acquisition sub-module is used for acquiring the historical click data of the first type of multimedia resources for each first type of multimedia resources;
the time attenuation coefficient training submodule is used for training the time attenuation coefficient of the first type of multimedia resources according to the historical click data of the first type of multimedia resources;
and the first click rate pre-estimation value calculation submodule is used for calculating the click rate pre-estimation value of the first type of multimedia resources according to the historical click data of the first type of multimedia resources and the time attenuation coefficient.
For the above apparatus, in a possible implementation manner, the time attenuation coefficient training sub-module is specifically configured to:
training a time attenuation coefficient theta of a first-class multimedia resource i by adopting a formula 1 according to historical click data of the first-class multimedia resource i, wherein the historical click data of the first-class multimedia resource i comprises the click quantity of the first-class multimedia resource i before a specified date;
f(t)＝Vi0×eθtformula 1;
wherein, Vi0Representing the click rate of the first type multimedia resources i on the specified date; t (0) represents the specified date, and f (0) represents Vi0(ii) a When t > 0, t represents the t-th day before the specified date, f (t)Representing the click rate of the first-class multimedia resources i on the t day before the specified date; when t is less than 0, t represents the day-tth after the specified day, f (t) represents the click rate estimated value of the first-class multimedia resources i on the day-tth after the specified day;
the first click rate pre-estimation value calculation submodule is specifically configured to:
and calculating the click rate estimated value of the first-class multimedia resource i on the-t day after the specified date by adopting the trained formula 1.
For the above apparatus, in a possible implementation manner, the time attenuation coefficient training sub-module is specifically configured to:
training a time attenuation coefficient theta of a first-class multimedia resource i by adopting a formula 2 according to historical click data of the first-class multimedia resource i, wherein the historical click data of the first-class multimedia resource i comprises the click quantity of the first-class multimedia resource i before a specified date;
g(t)＝lg(Vi0×eθt) Formula 2;
wherein, Vi0Representing the click rate of the first type multimedia resources i on the specified date; t-0 indicates the specified date, and g (0) -lgVi0(ii) a When t > 0, t represents the t day before the specified date, g (t) represents the logarithmic value of the click rate of the first-class multimedia resources i on the t day before the specified date; when t is less than 0, t represents the day-tth after the specified date, g (t) represents the logarithm value of the click rate estimated value of the first-class multimedia resources i on the day-tth after the specified date;
the first click rate pre-estimation value calculation submodule is specifically configured to:
and calculating the logarithm value of the click rate estimated value of the first-class multimedia resource i on the-t day after the specified date by adopting the trained formula 2.
For the above apparatus, in a possible implementation manner, the second click quantity estimated value determining module includes:
the matching sub-module is used for searching the first class of multimedia resources matched with the second class of multimedia resources according to the specified information of the second class of multimedia resources for each second class of multimedia resources; the specifying information includes at least one of: uploading time, time length and uploader information;
and the matching degree calculation operator module is used for calculating the matching degree of the second type of multimedia resources and the matched first type of multimedia resources according to the specified information of the second type of multimedia resources and the specified information of the matched first type of multimedia resources.
For the above apparatus, in a possible implementation manner, the second click quantity estimated value determining module includes:
and the second click rate estimated value calculation sub-module is used for taking the product of the matching degree and the click rate estimated value of the matched first-class multimedia resource as the click rate estimated value of the second-class multimedia resource.
For the apparatus, in a possible implementation manner, the recommending module includes:
the sequencing submodule is used for sequencing all the multimedia resources to be recommended according to the sequence of the click rate estimated values of the multimedia resources to be recommended from high to low;
and the recommending submodule is used for taking out M multimedia resources to be recommended from all the sequenced multimedia resources to be recommended to recommend, wherein the M is preset in size.
Advantageous effects
The method and the device for recommending the multimedia resources can ensure that newly uploaded multimedia resources can be also recommended at an opportunity, thereby improving the recommendation accuracy of the multimedia resources and improving the recommendation effect of the multimedia resources.
Other features and aspects of the present invention will become apparent from the following detailed description of exemplary embodiments, which proceeds with reference to the accompanying drawings.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate exemplary embodiments, features, and aspects of the invention and, together with the description, serve to explain the principles of the invention.
Fig. 1 shows a flowchart of an implementation of a recommendation method of multimedia resources according to an embodiment of the invention;
fig. 2 shows a flowchart of an exemplary specific implementation of step S101 of a recommendation method for multimedia resources according to an embodiment of the present invention;
fig. 3 shows a flowchart of another exemplary specific implementation of step S101 of a recommendation method for multimedia resources according to an embodiment of the present invention;
fig. 4 is a flowchart illustrating an exemplary specific implementation of step S102 of a recommendation method for multimedia resources according to an embodiment of the present invention;
fig. 5 is a flowchart illustrating an exemplary specific implementation of the method for recommending multimedia resources according to an embodiment of the present invention, in step S103, for each second-class multimedia resource, searching for a first-class multimedia resource matching the second-class multimedia resource, and calculating a matching degree between the second-class multimedia resource and the matched first-class multimedia resource;
FIG. 6 is a flowchart illustrating an exemplary implementation of step S104 of the method for recommending multimedia resources according to an embodiment of the present invention;
fig. 7 is a block diagram showing a construction of a recommendation apparatus for multimedia resources according to another embodiment of the present invention;
fig. 8 is a block diagram showing an exemplary configuration of a recommendation apparatus for multimedia resources according to another embodiment of the present invention;
fig. 9 is a block diagram showing a configuration of a recommendation apparatus for a multimedia asset according to another embodiment of the present invention.
Detailed Description
Various exemplary embodiments, features and aspects of the present invention will be described in detail below with reference to the accompanying drawings. In the drawings, like reference numbers can indicate functionally identical or similar elements. While the various aspects of the embodiments are presented in drawings, the drawings are not necessarily drawn to scale unless specifically indicated.
The word "exemplary" is used exclusively herein to mean "serving as an example, embodiment, or illustration. Any embodiment described herein as "exemplary" is not necessarily to be construed as preferred or advantageous over other embodiments.
Furthermore, in the following detailed description, numerous specific details are set forth in order to provide a better understanding of the present invention. It will be understood by those skilled in the art that the present invention may be practiced without some of these specific details. In some instances, methods, procedures, components, and circuits that are well known to those skilled in the art have not been described in detail so as not to obscure the present invention.
Example 1
Fig. 1 shows a flowchart of an implementation of a recommendation method for multimedia resources according to an embodiment of the present invention. As shown in fig. 1, the method mainly includes:
in step S101, multimedia resources to be recommended are divided into a first type of multimedia resources and a second type of multimedia resources.
In embodiments of the present invention, multimedia may be a composite of multiple media, including text, sound, and images, for example. For example, the multimedia resource according to the embodiment of the present invention may be a video, and is not limited herein.
As an example of the embodiment of the present invention, before step S101, the method further includes: and screening out the multimedia resources to be recommended from the multimedia resource library. For example, the multimedia resource to be recommended may be a multimedia resource that does not contain sensitive information in the multimedia resource library. The multimedia resource to be recommended may also be screened from the multimedia resource library according to other screening requirements, which is not limited herein.
In the embodiment of the invention, all multimedia resources to be recommended can be divided into the first type of multimedia resources and the second type of multimedia resources, and the click rate estimated values of the first type of multimedia resources and the second type of multimedia resources are calculated by adopting different methods respectively.
In step S102, for each first-class multimedia resource, a click volume estimation value of the first-class multimedia resource is determined according to historical click data of the first-class multimedia resource.
In step S103, for each second-class multimedia resource, a first-class multimedia resource matching the second-class multimedia resource is searched, a matching degree between the second-class multimedia resource and the matched first-class multimedia resource is calculated, and a click rate estimate of the second-class multimedia resource is calculated according to the matching degree and the click rate estimate of the matched first-class multimedia resource.
In the embodiment of the invention, for each first-class multimedia resource, the click rate estimated value of the first-class multimedia resource is determined according to the historical click data of the first-class multimedia resource. And for each second-class multimedia resource, searching the first-class multimedia resource matched with the second-class multimedia resource respectively. For example, for a second class of multimedia assets v1If the second type multimedia resource v is found1Matching multimedia assets of the first type v2Then calculate the second kind of multimedia resource v1With multimedia resources of the first type v2According to the matching degree and the first-class multimedia resource v2Calculating the second kind multimedia resource v by the click rate estimated value1Is estimated.
In step S104, the multimedia resource to be recommended is recommended according to the estimated click rate value of the multimedia resource to be recommended.
The multimedia resources to be recommended comprise first-class multimedia resources and second-class multimedia resources.
In a possible implementation manner, the estimated click rate value of the multimedia resource to be recommended and the application scenario may be combined to recommend the multimedia resource to be recommended. The application scenario may include application platform information and channel information, among others. For example, the Application platform may include a web page side and an APP (Application) side. According to different application platforms, the number of the multimedia resources to be recommended for recommendation may be different. The channels may include news channels, art channels, sports channels, and the like. The multimedia resources to be recommended for recommendation may also be different according to the channel.
Fig. 2 shows a flowchart of an exemplary specific implementation of step S101 of a method for recommending multimedia resources according to an embodiment of the present invention. As shown in fig. 2, dividing the multimedia resources to be recommended into a first class of multimedia resources and a second class of multimedia resources includes:
in step S201, historical click data of each multimedia resource to be recommended is acquired.
In step S202, for each multimedia resource to be recommended, whether the current click rate of the multimedia resource to be recommended is in a decline period is determined according to the historical click data of the multimedia resource to be recommended, if so, the multimedia resource to be recommended is determined as a first type of multimedia resource, otherwise, the multimedia resource to be recommended is determined as a second type of multimedia resource.
The fact that the current click rate of the multimedia resource to be recommended is in the decline period can mean that the click rate of the multimedia resource to be recommended shows a trend of decline along with the lapse of time.
Fig. 3 shows a flowchart of another exemplary specific implementation of step S101 of the method for recommending multimedia resources according to an embodiment of the present invention. As shown in fig. 3, dividing the multimedia resources to be recommended into a first class of multimedia resources and a second class of multimedia resources includes:
in step S301, the uploading time of each multimedia resource to be recommended is obtained.
In step S302, for each multimedia resource to be recommended, it is determined whether a time length between an uploading time of the multimedia resource to be recommended and a current system time is greater than a first preset value, if so, the multimedia resource to be recommended is determined as a first-class multimedia resource, otherwise, the multimedia resource to be recommended is determined as a second-class multimedia resource.
For example, the first preset value may be 48 hours, which is not limited herein.
Fig. 4 shows a flowchart of an exemplary specific implementation of step S102 of the method for recommending multimedia resources according to an embodiment of the present invention. As shown in fig. 4, for each first-class multimedia resource, determining a click rate pre-estimated value of the first-class multimedia resource according to historical click data of the first-class multimedia resource, includes:
in step S401, for each first-class multimedia resource, historical click data of the first-class multimedia resource is obtained.
In step S402, a time attenuation coefficient of the first type of multimedia resource is trained according to the historical click data of the first type of multimedia resource.
In step S403, a click volume pre-estimated value of the first type of multimedia resource is calculated according to the historical click data and the time decay coefficient of the first type of multimedia resource.
In one possible implementation manner, training the time attenuation coefficient of the first type of multimedia resource according to the historical click data of the first type of multimedia resource includes: training a time attenuation coefficient theta of a first-class multimedia resource i by adopting a formula 1 according to historical click data of the first-class multimedia resource i, wherein the historical click data of the first-class multimedia resource i comprises the click quantity of the first-class multimedia resource i before a specified date;
f(t)＝Vi0×eθtformula 1;
wherein, Vi0The click quantity of the first type of multimedia resources i on the specified date is represented; t is 0 indicating a specified date, f (0) is Vi0(ii) a When t is more than 0, t represents the tth day before the specified date, f (t) represents the click rate of the first multimedia resource i on the tth day before the specified date; when t is less than 0, t represents the day-tth after the specified day, and f (t) represents the click rate estimated value of the first-class multimedia resources i after the specified day-tth.
For example, the historical click data of the first type multimedia resource i comprises the click volume of the first type multimedia resource i on the specified date and the click volume of the first type multimedia resource i 15 days before the specified date. t ═ 1 represents the day before the specified date, f (1) represents the click rate of the first-class multimedia resource i on the day before the specified date, t ═ 2 represents the two days before the specified date, f (2) represents the click rate of the first-class multimedia resource i on the two days before the specified date, and so on, t ═ 15 represents the 15 days before the specified date, and f (15) represents the click rate of the first-class multimedia resource i on the 15 days before the specified date. When the time attenuation coefficient theta of the first type multimedia resource i is trained by adopting the formula 1, f (t), V in the formula 1 are more than 0i0And t is known according to f (t), Vi0And t can be trained to obtain the time attenuation coefficient theta of the first type multimedia resource i. The time attenuation coefficient theta can be corrected according to the historical click data of the first type multimedia resources i.
Calculating the click rate estimated value of the first type of multimedia resources according to the historical click data and the time attenuation coefficient of the first type of multimedia resources, wherein the method comprises the following steps: and calculating the click rate estimated value of the first-class multimedia resource i on the-t day after the specified date by adopting the trained formula 1.
t-1 represents the day before the specified date, and f (-1) represents the click volume of the first-type multimedia asset i on the day before the specified date. After the time attenuation coefficient theta of the first-class multimedia resource i is obtained through training, for t less than 0, V in formula 1i0θ and t are known so that an estimated click volume value can be calculated for a certain day after a given day. For example, the trained formula 1 is used to calculate the estimated click rate value of the first type multimedia resource i on the-t day after the specified day, which may be: and f (-1) is calculated by adopting the trained formula 1, and the f (-1) is determined as the click rate estimated value of the first-class multimedia resource i.
In one possible implementation manner, training the time attenuation coefficient of the first type of multimedia resource according to the historical click data of the first type of multimedia resource includes: training a time attenuation coefficient theta of a first-class multimedia resource i by adopting a formula 2 according to historical click data of the first-class multimedia resource i, wherein the historical click data of the first-class multimedia resource i comprises the click quantity of the first-class multimedia resource i before a specified date;
g(t)＝lg(Vi0×eθt) Formula 2;
wherein, Vi0The click quantity of the first type of multimedia resources i on the specified date is represented; t-0 indicates a specified date, and g (0) lgVi0(ii) a When t is more than 0, t represents the tth day before the appointed date, g (t) represents the logarithmic value of the click rate of the first multimedia resource i on the tth day before the appointed date; and when t is less than 0, t represents the day-tth after the specified date, and g (t) represents the logarithm value of the click rate estimated value of the first-class multimedia resources i after the specified date-tth.
Calculating the click rate estimated value of the first type of multimedia resources according to the historical click data and the time attenuation coefficient of the first type of multimedia resources, wherein the method comprises the following steps: and calculating the logarithm value of the click rate estimated value of the first-class multimedia resource i on the t day after the specified date by adopting the trained formula 2.
In this example, the clicks are logarithmized, thereby reducing the value for ease of computation and storage. In this example, the logarithmic value of the estimated click rate value of the first-class multimedia resource i on the t-th day after the specified date, which is calculated according to equation 2, can be used as the rank value of the first-class multimedia resource i. By calculating the rank value of the newly uploaded multimedia resources, more newly uploaded high-quality multimedia resources can be exposed.
Fig. 5 is a flowchart illustrating an exemplary specific implementation of the method for recommending multimedia resources according to an embodiment of the present invention, in step S103, for each second-class multimedia resource, searching for a first-class multimedia resource matching the second-class multimedia resource, and calculating a matching degree between the second-class multimedia resource and the matched first-class multimedia resource. As shown in fig. 5, for each second-class multimedia resource, searching for a first-class multimedia resource matching the second-class multimedia resource, and calculating a matching degree between the second-class multimedia resource and the matched first-class multimedia resource, the method includes:
in step S501, for each second-class multimedia resource, a first-class multimedia resource matching the second-class multimedia resource is searched according to the specified information of the second-class multimedia resource; the specifying information includes at least one of: upload time, length of time, and upload people information.
In step S502, a matching degree between the second type of multimedia resource and the matched first type of multimedia resource is calculated according to the specifying information of the second type of multimedia resource and the matching specifying information of the matched first type of multimedia resource.
And the matching degree of the second-class multimedia resources and the matched first-class multimedia resources is more than 0 and less than or equal to 1.
As an example of the embodiment of the present invention, the closer the uploading time of the second type of multimedia resource and the first type of multimedia resource is, the higher the matching degree between the two is.
As an example of the embodiment of the present invention, the closer the time length of the second type of multimedia resource and the first type of multimedia resource, the higher the matching degree between the two. The time length refers to a time length of a multimedia resource to be recommended, for example, if the multimedia resource to be recommended is a video, the time length of the multimedia resource to be recommended may refer to a video length.
As an example of the embodiment of the present invention, if the second type of multimedia resource is the same as the uploader of the first type of multimedia resource, the matching degree between the second type of multimedia resource and the first type of multimedia resource is higher.
As an example of the embodiment of the present invention, the matching degree between the second type of multimedia resource and the matched first type of multimedia resource is determined according to the uploading time, the time length and the uploader information, wherein the weight of the uploading time to the matching degree is λ1The weight of the time length to the matching degree is lambda2The weight of the uploader information to the matching degree is lambda3。
In a possible implementation manner, the calculating the click rate estimated value of the second type of multimedia resource according to the matching degree and the click rate estimated value of the matched first type of multimedia resource specifically includes: and taking the product of the matching degree and the click rate estimated value of the matched first-class multimedia resource as the click rate estimated value of the second-class multimedia resource.
Fig. 6 shows a flowchart of an exemplary specific implementation of step S104 of the method for recommending multimedia resources according to an embodiment of the present invention. As shown in fig. 6, recommending the multimedia resource to be recommended according to the estimated click rate value of the multimedia resource to be recommended includes:
in step S601, all the multimedia resources to be recommended are sorted according to the order from high to low of the estimated click rate of the multimedia resources to be recommended.
In step S602, M multimedia resources to be recommended that are ranked first are taken out from all the sorted multimedia resources to be recommended for recommendation, where the size of M is preset.
As an example of the embodiment of the present invention, recommending a multimedia resource to be recommended according to a click rate estimated value of the multimedia resource to be recommended includes: calculating by formula 1 to obtain a click rate estimated value of the multimedia resource to be recommended, sorting all the multimedia resources to be recommended according to the order of the click rate estimated value of the multimedia resource to be recommended from high to low, and then taking out M multimedia resources to be recommended from the sorted multimedia resources to be recommended to recommend.
As another example of the embodiment of the present invention, recommending a multimedia resource to be recommended according to a click rate estimated value of the multimedia resource to be recommended includes: and (3) calculating a logarithm value of the click rate estimated value of the multimedia resource to be recommended according to the formula 2, sequencing all the multimedia resources to be recommended according to the sequence of the logarithm value of the click rate estimated value of the multimedia resource to be recommended from high to low, and then taking out M multimedia resources to be recommended from the sequenced multimedia resources to be recommended for recommendation.
Therefore, the multimedia resources to be recommended are divided into the first type of multimedia resources and the second type of multimedia resources, the click rate pre-estimated value of the first type of multimedia resources is determined according to historical click data of the first type of multimedia resources, the click rate pre-estimated value of the second type of multimedia resources is calculated according to the click rate pre-estimated value of the first type of multimedia resources matched with the second type of multimedia resources and the matching degree between the first type of multimedia resources and the second type of multimedia resources, and the multimedia resources to be recommended are recommended according to the click rate pre-estimated value of the multimedia resources to be recommended.
Example 2
Fig. 7 is a block diagram illustrating a structure of a recommendation apparatus for multimedia resources according to another embodiment of the present invention. The apparatus may be used to execute the method for recommending multimedia resources shown in fig. 1. For convenience of explanation, only portions related to the embodiment of the present invention are shown in fig. 7.
As shown in fig. 7, the apparatus includes: a dividing module 71, configured to divide the multimedia resources to be recommended into a first class of multimedia resources and a second class of multimedia resources; a first click rate pre-estimated value determining module 72, configured to determine, for each first type of multimedia resource, a click rate pre-estimated value of the first type of multimedia resource according to historical click data of the first type of multimedia resource; a second click rate pre-estimated value determining module 73, configured to, for each second-class multimedia resource, search for a first-class multimedia resource that matches the second-class multimedia resource, calculate a matching degree between the second-class multimedia resource and the matched first-class multimedia resource, and calculate a click rate pre-estimated value of the second-class multimedia resource according to the matching degree and the click rate pre-estimated value of the matched first-class multimedia resource; and the recommending module 74 is configured to recommend the multimedia resource to be recommended according to the estimated click rate value of the multimedia resource to be recommended.
Fig. 8 is a block diagram illustrating an exemplary configuration of a recommendation apparatus for multimedia resources according to another embodiment of the present invention. The apparatus may be used to execute the method for recommending multimedia resources shown in fig. 1 to 6. For convenience of explanation, only a part related to the present example is shown in fig. 8. Components in fig. 8 that are numbered the same as those in fig. 7 have the same functions, and detailed descriptions of these components are omitted for the sake of brevity. As shown in fig. 8:
in one possible implementation, the dividing module 71 includes: the historical click data acquisition submodule 711 is configured to respectively acquire historical click data of each multimedia resource to be recommended; the first scheduling module 712 is configured to, for each multimedia resource to be recommended, determine whether a current click rate of the multimedia resource to be recommended is in a decline period according to historical click data of the multimedia resource to be recommended, determine, if yes, the multimedia resource to be recommended as a first type of multimedia resource, and otherwise, determine, as a second type of multimedia resource, the multimedia resource to be recommended.
In one possible implementation, the dividing module 71 includes: an upload time obtaining submodule 713, configured to obtain upload time of each multimedia resource to be recommended respectively; the second partitioning submodule 714 is configured to, for each multimedia resource to be recommended, determine whether a time length between an uploading time of the multimedia resource to be recommended and a current system time is greater than a first preset value, determine, if yes, the multimedia resource to be recommended as a first class of multimedia resource, and otherwise, determine, as a second class of multimedia resource, the multimedia resource to be recommended.
In one possible implementation, the first estimated click quantity determining module 72 includes: a historical click data obtaining sub-module 721, configured to, for each first type of multimedia resource, obtain historical click data of the first type of multimedia resource; the time attenuation coefficient training submodule 722 is used for training the time attenuation coefficient of the first-class multimedia resource according to the historical click data of the first-class multimedia resource; and the first click rate pre-estimation value calculation sub-module 723 is configured to calculate the click rate pre-estimation value of the first type of multimedia resource according to the historical click data of the first type of multimedia resource and the time attenuation coefficient.
In one possible implementation, the time attenuation coefficient training submodule 722 is specifically configured to:
training a time attenuation coefficient theta of a first-class multimedia resource i by adopting a formula 1 according to historical click data of the first-class multimedia resource i, wherein the historical click data of the first-class multimedia resource i comprises the click quantity of the first-class multimedia resource i before a specified date;
f(t)＝Vi0×eθtformula 1;
wherein, Vi0Representing the click rate of the first type multimedia resources i on the specified date; t (0) represents the specified date, and f (0) represents Vi0(ii) a When t > 0, t represents the tth day before the specified date, f (t) represents the click rate of the first-class multimedia resources i on the tth day before the specified date; when t is less than 0, t represents the day-tth after the specified day, f (t) represents the click rate estimated value of the first-class multimedia resources i on the day-tth after the specified day;
the first click quantity pre-estimation value calculation submodule 723 is specifically configured to:
and calculating the click rate estimated value of the first-class multimedia resource i on the-t day after the specified date by adopting the trained formula 1.
In one possible implementation, the time attenuation coefficient training submodule 722 is specifically configured to:
training a time attenuation coefficient theta of a first-class multimedia resource i by adopting a formula 2 according to historical click data of the first-class multimedia resource i, wherein the historical click data of the first-class multimedia resource i comprises the click quantity of the first-class multimedia resource i before a specified date;
g(t)＝lg(Vi0×eθt) Formula 2;
wherein, Vi0Representing the click rate of the first type multimedia resources i on the specified date; t-0 indicates the specified date, and g (0) -lgVi0(ii) a When t > 0, t represents the t-th day before the specified date, g (t) represents the t-th day before the specified date of the first type of multimedia resource iA logarithmic value of the click rate; when t is less than 0, t represents the day-tth after the specified date, g (t) represents the logarithm value of the click rate estimated value of the first-class multimedia resources i on the day-tth after the specified date;
the first click quantity pre-estimation value calculation submodule 723 is specifically configured to:
and calculating the logarithm value of the click rate estimated value of the first-class multimedia resource i on the-t day after the specified date by adopting the trained formula 2.
In a possible implementation manner, the second click quantity estimated value determining module 73 includes: a matching sub-module 731, configured to, for each second-class multimedia resource, search, according to the specified information of the second-class multimedia resource, a first-class multimedia resource that matches the second-class multimedia resource; the specifying information includes at least one of: uploading time, time length and uploader information; the matching degree operator module 732 is configured to calculate a matching degree between the second type of multimedia resource and the matched first type of multimedia resource according to the specific information of the second type of multimedia resource and the specific information of the matched first type of multimedia resource.
In a possible implementation manner, the second click quantity estimated value determining module 73 includes:
and the second click rate estimated value calculating sub-module 733, configured to use a product of the matching degree and the click rate estimated value of the matched first-class multimedia resource as the click rate estimated value of the second-class multimedia resource.
In one possible implementation, the recommendation module 74 includes: the ranking submodule 741 is configured to rank all the multimedia resources to be recommended according to the order from high to low of the click quantity estimated values of the multimedia resources to be recommended; and a recommending submodule 742, configured to take out the top M sequenced to-be-recommended multimedia resources from all the sequenced to-be-recommended multimedia resources for recommendation, where the size of M is preset.
It should be noted that, in this way, the multimedia resources to be recommended are divided into the first type of multimedia resources and the second type of multimedia resources, the click rate pre-estimated value of the first type of multimedia resources is determined according to the historical click data of the first type of multimedia resources, the click rate pre-estimated value of the second type of multimedia resources is calculated according to the click rate pre-estimated value of the first type of multimedia resources matched with the second type of multimedia resources and the matching degree between the click rate pre-estimated value and the click rate pre-estimated value, and then the multimedia resources to be recommended are recommended according to the click rate pre-estimated value of the multimedia resources to be recommended.
Example 3
Fig. 9 is a block diagram showing a configuration of a recommendation apparatus for a multimedia asset according to another embodiment of the present invention. The recommendation device 1100 for multimedia resources may be a host server with computing power, a personal computer PC, or a portable computer or terminal that can be carried, etc. The specific embodiments of the present invention do not limit the specific implementation of the compute node.
The recommendation device 1100 for multimedia resources includes a processor (processor)1110, a communication Interface (Communications Interface)1120, a memory (memory)1130, and a bus 1140. The processor 1110, the communication interface 1120, and the memory 1130 communicate with each other via the bus 1140.
The communication interface 1120 is used to communicate with network devices, including, for example, virtual machine management centers, shared storage, and the like.
Processor 1110 is configured to execute programs. Processor 1110 may be a central processing unit CPU, or an Application Specific Integrated Circuit (ASIC), or one or more Integrated circuits configured to implement embodiments of the present invention.
The memory 1130 is used to store files. The memory 1130 may comprise high-speed RAM memory, and may also include non-volatile memory (non-volatile memory), such as at least one disk memory. The memory 1130 may also be a memory array. The storage 1130 may also be partitioned and the blocks may be combined into virtual volumes according to certain rules.
In one possible embodiment, the program may be a program code including computer operation instructions. The procedure is particularly useful for: the operations of the steps in example 1 were carried out.
Those of ordinary skill in the art will appreciate that the various illustrative elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, or combinations of computer software and electronic hardware. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the solution. Those skilled in the art may select different ways to implement the described functionality for specific applications, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.
If the described functionality is implemented in the form of computer software and sold or used as a stand-alone product, it is to some extent possible to consider all or part of the technical solution of the invention (for example, the part contributing to the prior art) to be embodied in the form of a computer software product. The computer software product is generally stored in a non-volatile storage medium readable by a computer and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device) to perform all or part of the steps of the methods according to the embodiments of the present invention. The storage medium includes various media capable of storing program codes, such as a usb disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk, or an optical disk.
The above description is only for the specific embodiments of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art can easily conceive of the changes or substitutions within the technical scope of the present invention, and all the changes or substitutions should be covered within the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the appended claims.

Claims (14)

1. A method for recommending multimedia resources, comprising:
dividing multimedia resources to be recommended into first-class multimedia resources and second-class multimedia resources;
for each first-class multimedia resource, determining a click rate pre-estimated value of the first-class multimedia resource according to historical click data of the first-class multimedia resource;
for each second-class multimedia resource, searching a first-class multimedia resource matched with the second-class multimedia resource, calculating the matching degree of the second-class multimedia resource and the matched first-class multimedia resource, and calculating the click rate estimated value of the second-class multimedia resource according to the matching degree and the click rate estimated value of the matched first-class multimedia resource;
recommending the multimedia resource to be recommended according to the click rate estimated value of the multimedia resource to be recommended;
the method for dividing the multimedia resources to be recommended into the first-class multimedia resources and the second-class multimedia resources comprises the following steps:
respectively acquiring historical click data of each multimedia resource to be recommended; for each multimedia resource to be recommended, judging whether the current click rate of the multimedia resource to be recommended is in a decline period according to historical click data of the multimedia resource to be recommended, if so, determining the multimedia resource to be recommended as a first type of multimedia resource, otherwise, determining the multimedia resource to be recommended as a second type of multimedia resource;
alternatively, the first and second electrodes may be,
respectively acquiring the uploading time of each multimedia resource to be recommended; and for each multimedia resource to be recommended, judging whether the time length of the uploading time of the multimedia resource to be recommended from the current system time is greater than a first preset value, if so, determining the multimedia resource to be recommended as a first class of multimedia resource, and otherwise, determining the multimedia resource to be recommended as a second class of multimedia resource.
2. The method of claim 1, wherein for each of the first type multimedia resources, determining a click rate estimate for the first type multimedia resource based on historical click data for the first type multimedia resource comprises:
for each first-class multimedia resource, acquiring historical click data of the first-class multimedia resource;
training the time attenuation coefficient of the first type of multimedia resources according to the historical click data of the first type of multimedia resources;
and calculating the click rate pre-estimated value of the first type of multimedia resources according to the historical click data of the first type of multimedia resources and the time attenuation coefficient.
3. The method of claim 2, wherein training the time decay coefficient of the first type of multimedia asset according to the historical click data of the first type of multimedia asset comprises:
training a time attenuation coefficient theta of a first-class multimedia resource i by adopting a formula 1 according to historical click data of the first-class multimedia resource i, wherein the historical click data of the first-class multimedia resource i comprises the click quantity of the first-class multimedia resource i before a specified date;
f(t)＝Vi0×eθtformula 1;
wherein, Vi0Representing the click rate of the first type multimedia resources i on the specified date; t (0) represents the specified date, and f (0) represents Vi0(ii) a When t > 0, t represents the tth day before the specified date, f (t) represents the click rate of the first-class multimedia resources i on the tth day before the specified date; when t is less than 0, t represents the day-tth after the specified day, f (t) represents the click rate estimated value of the first-class multimedia resources i on the day-tth after the specified day;
calculating the click rate pre-estimated value of the first type of multimedia resources according to the historical click data of the first type of multimedia resources and the time attenuation coefficient, wherein the method comprises the following steps:
and calculating the click rate estimated value of the first-class multimedia resource i on the-t day after the specified date by adopting the trained formula 1.
4. The method of claim 2, wherein training the time decay coefficient of the first type of multimedia asset according to the historical click data of the first type of multimedia asset comprises:
training a time attenuation coefficient theta of a first-class multimedia resource i by adopting a formula 2 according to historical click data of the first-class multimedia resource i, wherein the historical click data of the first-class multimedia resource i comprises the click quantity of the first-class multimedia resource i before a specified date;
g(t)＝lg(Vi0×eθt) Formula 2;
wherein, Vi0Representing the click rate of the first type multimedia resources i on the specified date; t is 0 and g (0) is lg Vi0(ii) a When t > 0, t represents the t day before the specified date, g (t) represents the logarithmic value of the click rate of the first-class multimedia resources i on the t day before the specified date; when t is less than 0, t represents the day-tth after the specified date, g (t) represents the logarithm value of the click rate estimated value of the first-class multimedia resources i on the day-tth after the specified date;
calculating the click rate pre-estimated value of the first type of multimedia resources according to the historical click data of the first type of multimedia resources and the time attenuation coefficient, wherein the method comprises the following steps:
and calculating the logarithm value of the click rate estimated value of the first-class multimedia resource i on the-t day after the specified date by adopting the trained formula 2.
5. The method of claim 1, wherein for each of the second types of multimedia resources, finding a first type of multimedia resource matching the second type of multimedia resource, and calculating a matching degree between the second type of multimedia resource and the matched first type of multimedia resource comprises:
for each second-class multimedia resource, searching a first-class multimedia resource matched with the second-class multimedia resource according to the specified information of the second-class multimedia resource; the specifying information includes at least one of: uploading time, time length and uploader information;
and calculating the matching degree of the second type of multimedia resources and the matched first type of multimedia resources according to the specified information of the second type of multimedia resources and the specified information of the matched first type of multimedia resources.
6. The method according to any of claims 1 to 5, wherein the estimated click rate value of the second type of multimedia resource is calculated according to the matching degree and the estimated click rate value of the matched first type of multimedia resource, specifically:
and taking the product of the matching degree and the click rate estimated value of the matched first-class multimedia resource as the click rate estimated value of the second-class multimedia resource.
7. The method according to any one of claims 1 to 5, wherein recommending the multimedia resource to be recommended according to the click rate estimate of the multimedia resource to be recommended comprises:
sequencing all the multimedia resources to be recommended according to the sequence of the click rate estimated values of the multimedia resources to be recommended from high to low;
and taking out M multimedia resources to be recommended which are sequenced at the front from all the sequenced multimedia resources to be recommended for recommendation, wherein the size of M is preset.
8. An apparatus for recommending multimedia resources, comprising:
the system comprises a dividing module, a recommending module and a recommending module, wherein the dividing module is used for dividing the multimedia resources to be recommended into a first type of multimedia resources and a second type of multimedia resources;
the first click rate pre-estimated value determining module is used for determining the click rate pre-estimated value of each first type of multimedia resource according to historical click data of the first type of multimedia resource;
the second click rate pre-estimated value determining module is used for searching a first class of multimedia resources matched with the second class of multimedia resources for each second class of multimedia resources, calculating the matching degree of the second class of multimedia resources and the matched first class of multimedia resources, and calculating the click rate pre-estimated value of the second class of multimedia resources according to the matching degree and the click rate pre-estimated value of the matched first class of multimedia resources;
the recommending module is used for recommending the multimedia resource to be recommended according to the click rate estimated value of the multimedia resource to be recommended;
the dividing module comprises a historical click data acquisition sub-module and a first dividing sub-module, or the dividing module comprises an uploading time acquisition sub-module and a second dividing sub-module;
the historical click data acquisition submodule is used for respectively acquiring the historical click data of each multimedia resource to be recommended;
the first dividing module is used for judging whether the current click rate of the multimedia resources to be recommended is in a decline period or not according to the historical click data of the multimedia resources to be recommended for each multimedia resource to be recommended, if so, the multimedia resources to be recommended are determined as first-class multimedia resources, and if not, the multimedia resources to be recommended are determined as second-class multimedia resources;
the uploading time obtaining submodule is used for respectively obtaining the uploading time of each multimedia resource to be recommended;
the second division submodule is used for judging whether the time length of the uploading time of the multimedia resources to be recommended from the current system time is larger than a first preset value or not for each multimedia resource to be recommended, if so, the multimedia resources to be recommended are determined as first-class multimedia resources, and otherwise, the multimedia resources to be recommended are determined as second-class multimedia resources.
9. The apparatus of claim 8, wherein the first estimated click quantity determination module comprises:
the historical click data acquisition sub-module is used for acquiring the historical click data of the first type of multimedia resources for each first type of multimedia resources;
the time attenuation coefficient training submodule is used for training the time attenuation coefficient of the first type of multimedia resources according to the historical click data of the first type of multimedia resources;
and the first click rate pre-estimation value calculation submodule is used for calculating the click rate pre-estimation value of the first type of multimedia resources according to the historical click data of the first type of multimedia resources and the time attenuation coefficient.
10. The apparatus of claim 9, wherein the time decay factor training submodule is specifically configured to:
training a time attenuation coefficient theta of a first-class multimedia resource i by adopting a formula 1 according to historical click data of the first-class multimedia resource i, wherein the historical click data of the first-class multimedia resource i comprises the click quantity of the first-class multimedia resource i before a specified date;
f(t)＝Vi0×eθtformula 1;
wherein, Vi0Representing the click rate of the first type multimedia resources i on the specified date; t (0) represents the specified date, and f (0) represents Vi0(ii) a When t > 0, t represents the tth day before the specified date, f (t) represents the click rate of the first-class multimedia resources i on the tth day before the specified date; when t is less than 0, t represents the day-tth after the specified day, f (t) represents the click rate estimated value of the first-class multimedia resources i on the day-tth after the specified day;
the first click rate pre-estimation value calculation submodule is specifically configured to:
and calculating the click rate estimated value of the first-class multimedia resource i on the-t day after the specified date by adopting the trained formula 1.
11. The apparatus of claim 9, wherein the time decay factor training submodule is specifically configured to:
training a time attenuation coefficient theta of a first-class multimedia resource i by adopting a formula 2 according to historical click data of the first-class multimedia resource i, wherein the historical click data of the first-class multimedia resource i comprises the click quantity of the first-class multimedia resource i before a specified date;
g(t)＝lg(Vi0×eθt) Formula 2;
wherein, Vi0Representing the click rate of the first type multimedia resources i on the specified date; t is 0 and g (0) is lg Vi0(ii) a When t > 0, t represents the t day before the specified date, g (t) represents the logarithmic value of the click rate of the first-class multimedia resources i on the t day before the specified date; when t is less than 0, t represents the day-tth after the specified date, g (t) represents the logarithm value of the click rate estimated value of the first-class multimedia resources i on the day-tth after the specified date;
the first click rate pre-estimation value calculation submodule is specifically configured to:
and calculating the logarithm value of the click rate estimated value of the first-class multimedia resource i on the-t day after the specified date by adopting the trained formula 2.
12. The apparatus of claim 8, wherein the second click rate estimate determining module comprises:
the matching sub-module is used for searching the first class of multimedia resources matched with the second class of multimedia resources according to the specified information of the second class of multimedia resources for each second class of multimedia resources; the specifying information includes at least one of: uploading time, time length and uploader information;
and the matching degree calculation operator module is used for calculating the matching degree of the second type of multimedia resources and the matched first type of multimedia resources according to the specified information of the second type of multimedia resources and the specified information of the matched first type of multimedia resources.
13. The apparatus according to any one of claims 8 to 12, wherein the second click rate estimate determining module comprises:
and the second click rate estimated value calculation sub-module is used for taking the product of the matching degree and the click rate estimated value of the matched first-class multimedia resource as the click rate estimated value of the second-class multimedia resource.
14. The apparatus of any one of claims 8 to 12, wherein the recommendation module comprises:
the sequencing submodule is used for sequencing all the multimedia resources to be recommended according to the sequence of the click rate estimated values of the multimedia resources to be recommended from high to low;
and the recommending submodule is used for taking out M multimedia resources to be recommended from all the sequenced multimedia resources to be recommended to recommend, wherein the M is preset in size.
CN201610594958.7A 2016-07-26 2016-07-26 Multimedia resource recommendation method and device Expired - Fee Related CN106227834B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201610594958.7A CN106227834B (en) 2016-07-26 2016-07-26 Multimedia resource recommendation method and device

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201610594958.7A CN106227834B (en) 2016-07-26 2016-07-26 Multimedia resource recommendation method and device

Publications (2)

Publication Number Publication Date
CN106227834A CN106227834A (en) 2016-12-14
CN106227834B true CN106227834B (en) 2020-03-27

Family

ID=57534156

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201610594958.7A Expired - Fee Related CN106227834B (en) 2016-07-26 2016-07-26 Multimedia resource recommendation method and device

Country Status (1)

Country Link
CN (1) CN106227834B (en)

Families Citing this family (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108804439A (en) * 2017-04-26 2018-11-13 合信息技术（北京）有限公司 The recommendation method and device of multimedia resource
CN108804452B (en) * 2017-04-28 2021-06-04 阿里巴巴（中国）有限公司 Multimedia resource cover display method and device
CN109388739A (en) * 2017-08-03 2019-02-26 合信息技术（北京）有限公司 The recommended method and device of multimedia resource
CN107748787A (en) * 2017-10-27 2018-03-02 姜俊 A kind of heatrate processing method and processing device
CN109068158B (en) * 2018-09-18 2021-06-22 上海众引文化传播股份有限公司 Short video value estimation recommendation system based on network platform
CN111199412B (en) * 2018-11-19 2023-05-05 腾讯科技（深圳）有限公司 Information processing method, apparatus, computer device, and storage medium
CN110990690A (en) * 2019-11-12 2020-04-10 上海易点时空网络有限公司 Post recommendation method and device
CN110909176B (en) * 2019-11-20 2021-03-02 腾讯科技（深圳）有限公司 Data recommendation method and device, computer equipment and storage medium
CN112948402A (en) * 2021-01-15 2021-06-11 浙江大华技术股份有限公司 Database updating method and device, electronic equipment and computer readable storage medium
CN113487368A (en) * 2021-07-29 2021-10-08 宿迁市盛邦信息科技有限公司 Scientific and technological information push service system based on big data
CN116389444B (en) * 2023-04-10 2023-09-15 北京智享嘉网络信息技术有限公司 Traffic scheduling method and system based on user web application

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103164481A (en) * 2011-12-16 2013-06-19 盛乐信息技术（上海）有限公司 Recommendation method and system of video with largest rising trend
CN104520881A (en) * 2012-06-22 2015-04-15 谷歌公司 Ranking nearby destinations based on visit likelihoods and predicting future visits to places from location history
CN105184618A (en) * 2015-10-20 2015-12-23 广州唯品会信息科技有限公司 Commodity individual recommendation method for new users and system
CN105447045A (en) * 2014-09-02 2016-03-30 阿里巴巴集团控股有限公司 Information ordering method and device and information providing method and system
CN105760530A (en) * 2016-03-03 2016-07-13 北京百度网讯科技有限公司 Information card sequencing method and device based on artificial intelligence

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6904408B1 (en) * 2000-10-19 2005-06-07 Mccarthy John Bionet method, system and personalized web content manager responsive to browser viewers' psychological preferences, behavioral responses and physiological stress indicators

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103164481A (en) * 2011-12-16 2013-06-19 盛乐信息技术（上海）有限公司 Recommendation method and system of video with largest rising trend
CN104520881A (en) * 2012-06-22 2015-04-15 谷歌公司 Ranking nearby destinations based on visit likelihoods and predicting future visits to places from location history
CN105447045A (en) * 2014-09-02 2016-03-30 阿里巴巴集团控股有限公司 Information ordering method and device and information providing method and system
CN105184618A (en) * 2015-10-20 2015-12-23 广州唯品会信息科技有限公司 Commodity individual recommendation method for new users and system
CN105760530A (en) * 2016-03-03 2016-07-13 北京百度网讯科技有限公司 Information card sequencing method and device based on artificial intelligence

Also Published As

Publication number Publication date
CN106227834A (en) 2016-12-14

Similar Documents

Publication Publication Date Title
CN106227834B (en) Multimedia resource recommendation method and device
US10405016B2 (en) Recommending media items based on take rate signals
CN106326391B (en) Multimedia resource recommendation method and device
US9467744B2 (en) Comment-based media classification
CN108345702A (en) Entity recommends method and apparatus
US8583685B2 (en) Determination of category information using multiple stages
CN108108821A (en) Model training method and device
CN110896488B (en) Recommendation method for live broadcast room and related equipment
CN108062573A (en) Model training method and device
WO2015081915A1 (en) File recommendation method and device
WO2012092196A1 (en) Recommendation of search keywords based on indication of user intention
CN103377232A (en) Headline keyword recommendation method and system
CN105915956A (en) Video content recommendation method, device, server and system
CN112423134B (en) Video content recommendation method and device, computer equipment and storage medium
CN106897398B (en) Video display method and device
CN106354867A (en) Multimedia resource recommendation method and device
CN113873330B (en) Video recommendation method and device, computer equipment and storage medium
CN106202474A (en) A kind of object recommendation method and apparatus
CN110704677A (en) Program recommendation method and device, readable storage medium and terminal equipment
CN105022807A (en) Information recommendation method and apparatus
CN113239182A (en) Article recommendation method and device, computer equipment and storage medium
CN111078997B (en) Information recommendation method and device
CN111241381A (en) Information recommendation method and device, electronic equipment and computer-readable storage medium
US10362349B1 (en) Detecting channel similarity based on content reuse
CN115544299A (en) Video recommendation method and device

Legal Events

Date Code Title Description
C06 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
CP03 Change of name, title or address

Address after: 100080 Beijing Haidian District city Haidian street A Sinosteel International Plaza No. 8 block 5 layer A, C

Patentee after: Youku network technology (Beijing) Co.,Ltd.

Address before: 100080 area a and C, 5 / F, block a, Sinosteel International Plaza, No. 8, Haidian Street, Haidian District, Beijing

Patentee before: 1VERGE INTERNET TECHNOLOGY (BEIJING) Co.,Ltd.

CP03 Change of name, title or address
TR01 Transfer of patent right

Effective date of registration: 20200520

Address after: 310052 room 508, floor 5, building 4, No. 699, Wangshang Road, Changhe street, Binjiang District, Hangzhou City, Zhejiang Province

Patentee after: Alibaba (China) Co.,Ltd.

Address before: 100080 Beijing Haidian District city Haidian street A Sinosteel International Plaza No. 8 block 5 layer A, C

Patentee before: Youku network technology (Beijing) Co.,Ltd.

TR01 Transfer of patent right
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20200327

Termination date: 20200726

CF01 Termination of patent right due to non-payment of annual fee