CN116521974A - Media content recommendation method and device, electronic equipment and readable storage medium - Google Patents

Media content recommendation method and device, electronic equipment and readable storage medium Download PDF

Info

Publication number
CN116521974A
CN116521974A CN202210066809.9A CN202210066809A CN116521974A CN 116521974 A CN116521974 A CN 116521974A CN 202210066809 A CN202210066809 A CN 202210066809A CN 116521974 A CN116521974 A CN 116521974A
Authority
CN
China
Prior art keywords
media content
feedback information
preset
target media
information
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202210066809.9A
Other languages
Chinese (zh)
Inventor
李倩
孔维莲
冯俊兰
邓超
曾海涛
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.)
China Mobile Communications Group Co Ltd
China Mobile Communications Ltd Research Institute
Original Assignee
China Mobile Communications Group Co Ltd
China Mobile Communications Ltd Research Institute
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 China Mobile Communications Group Co Ltd, China Mobile Communications Ltd Research Institute filed Critical China Mobile Communications Group Co Ltd
Priority to CN202210066809.9A priority Critical patent/CN116521974A/en
Publication of CN116521974A publication Critical patent/CN116521974A/en
Pending legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90—Details of database functions independent of the retrieved data types
    • G06F16/95—Retrieval from the web
    • G06F16/953—Querying, e.g. by the use of web search engines
    • G06F16/9535—Search customisation based on user profiles and personalisation

Abstract

The application provides a media content recommendation method, a device, an electronic device and a readable storage medium, wherein the method comprises the following steps: acquiring feedback information of a user on target media content; under the condition that the feedback information is not matched with preset feedback information, determining a correction coefficient of the target media content according to the feedback information and the preset feedback information, wherein the preset feedback information is associated with display information of the target media content at a client, and the display information comprises display position information and/or display time information; and correcting the recommended parameters of the target media content according to the correction coefficient. The method and the device can improve flexibility of media content recommendation.

Description

Media content recommendation method and device, electronic equipment and readable storage medium
Technical Field
The embodiment of the application relates to the technical field of content recommendation, in particular to a media content recommendation method, a device, electronic equipment and a readable storage medium.
Background
In recent years, with the rapid development of computer technology and internet technology, the variety of media content is more and more varied, and how to reasonably recommend media content to users has become an important research direction.
Currently, to attract users, and increase the viscosity of users, each media website typically determines a batch of media content to recommend to all users for increasing the exposure of the media content based on a combination of events or market demands. However, there may be content in the media content that is not of interest to the user, which would occupy unnecessary network resources and waste traffic from some exposure bits.
Disclosure of Invention
An object of the embodiments of the present application is to provide a media content recommendation method, apparatus, electronic device, and readable storage medium, which solve the problem of recommended resource waste in the prior art.
In order to solve the above problems, in a first aspect, an embodiment of the present application provides a media content recommendation method, including:
acquiring feedback information of a user on target media content;
under the condition that the feedback information is not matched with preset feedback information, determining a correction coefficient of the target media content according to the feedback information and the preset feedback information, wherein the preset feedback information is associated with display information of the target media content at a client, and the display information comprises display position information and/or display time information;
And correcting the recommended parameters of the target media content according to the correction coefficient.
Optionally, before the obtaining the feedback information of the user on the target media content, the method further includes:
determining a target user based on content tag information of target media content, wherein the user tag information of the target user is matched with the content tag information of the target media content;
the obtaining feedback information of the user on the target media content comprises the following steps:
and acquiring feedback information of the target user on the target media content.
Optionally, the determining, when the feedback information does not match with the preset feedback information, a correction coefficient of the target media content according to the feedback information and the preset feedback information includes:
acquiring first feedback information of a user on the target media content in a first time period;
under the condition that the first feedback information is not matched with first preset feedback information, determining a first correction coefficient of the target media content according to the first feedback information and the first preset feedback information, wherein the first preset feedback information is preset feedback information corresponding to the first time period;
And correcting the recommended parameters of the target media content according to the correction coefficient, wherein the method comprises the following steps:
and correcting the recommended parameters of the target media content in a second time period according to the first correction coefficient, wherein the second time period is continuous with the first time period and is after the first time period.
Optionally, the feedback information includes a click rate, and the preset feedback information includes a preset click rate;
and under the condition that the feedback information is not matched with preset feedback information, determining the correction coefficient of the target media content according to the feedback information and the preset feedback information, including:
and under the condition that the click rate is not matched with the preset click rate, determining a correction coefficient of the target media content according to the ratio of the click rate to the preset click rate.
Optionally, the preset click rate is determined according to the following formula:
click j ＝exp(a·j b +c)
wherein click j J is the display order, a, b and c of the target media content at the client for the preset click quantityAnd the value is determined according to the display time of the target media content at the client.
Optionally, the correcting the recommended parameters of the target media content according to the correction coefficient includes:
Acquiring a first recommendation score of the target media content before correction;
and determining a second recommendation score of the target media content after correction according to the product of the correction coefficient and the first recommendation score.
Optionally, before the obtaining the feedback information of the user on the target media content, the method further includes:
determining a first media content;
recall the first media content based on a preset recall policy;
and determining the target media content in the first media content based on a preset screening strategy.
Optionally, determining the first media content includes:
determining candidate media content;
determining estimated flow information of each candidate media content according to the characteristic information of each candidate media content;
and determining first media content in the candidate media content according to the estimated flow information of each candidate media content.
Optionally, recall the first media content based on a preset recall policy, including:
and recalling the first media content according to the identification information of the first media content.
In a second aspect, an embodiment of the present application provides a media content recommendation device, including:
The first acquisition module is used for acquiring feedback information of a user on target media content;
the first determining module is used for determining a correction coefficient of the target media content according to the feedback information and the preset feedback information when the feedback information is not matched with the preset feedback information, wherein the preset feedback information is associated with display information of the target media content at a client, and the display information comprises display position information and/or display time information;
and the correction module is used for correcting the recommended parameters of the target media content according to the correction coefficient.
Optionally, the apparatus further comprises:
the second determining module is used for determining a target user based on content tag information of target media content, wherein the user tag information of the target user is matched with the content tag information of the target media content;
the first acquisition module is used for:
and acquiring feedback information of the target user on the target media content.
Optionally, the first determining module includes:
the first acquisition unit is used for acquiring first feedback information of the user on the target media content in a first time period;
the first determining unit is configured to determine, according to the first feedback information and the first preset feedback information, a first correction coefficient of the target media content when the first feedback information is not matched with the first preset feedback information, where the first preset feedback information is preset feedback information corresponding to the first time period;
The correction module is used for:
and correcting the recommended parameters of the target media content in a second time period according to the first correction coefficient, wherein the second time period is continuous with the first time period and is after the first time period.
Optionally, the feedback information includes a click rate, and the preset feedback information includes a preset click rate;
the first determining module is used for:
and under the condition that the click rate is not matched with the preset click rate, determining a correction coefficient of the target media content according to the ratio of the click rate to the preset click rate.
Optionally, the preset click rate is determined according to the following formula:
click j ＝exp(a·j b +c)
wherein click j And j is the display order of the target media content at the client for the preset click quantity, and the values of a, b and c are determined according to the display time of the target media content at the client.
Optionally, the correction module includes:
a second obtaining unit, configured to obtain a first recommendation score of the target media content before correction;
and the second determining unit is used for determining a second recommendation score of the target media content after correction according to the product of the correction coefficient and the first recommendation score.
Optionally, the apparatus further comprises:
a third determining module for determining the first media content;
a recall module for recalling the first media content based on a preset recall policy;
and the screening module is used for determining the target media content in the first media content based on a preset screening strategy.
Optionally, determining the third determining module includes:
a fourth determining unit configured to determine candidate media contents;
a fifth determining unit, configured to determine estimated traffic information of each candidate media content according to feature information of each candidate media content;
and a sixth determining unit, configured to determine a first media content in the candidate media contents according to the estimated traffic information of each candidate media content.
Optionally, the recall module is configured to:
and recalling the first media content according to the identification information of the first media content.
In a third aspect, embodiments of the present application provide an electronic device comprising a processor, a memory, and a computer program stored on the memory and executable on the processor, the computer program implementing the steps of the media content recommendation method as described above when executed by the processor.
In a fourth aspect, embodiments of the present application provide a computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements the steps of a media content recommendation method as described above.
In the embodiment of the application, after recommending the target media content to the client, the server can correct the recommendation parameters of the target media content by acquiring the feedback information of the user on the target media content, so that the target media content can be sufficiently exposed and promoted, has the recommendation parameters matched with the interesting degree of the user, and avoids occupying unnecessary recommendation resources, and improves the flexibility of media content recommendation.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are needed in the description of the embodiments of the present application will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present application, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is one of the flowcharts of a media content recommendation method provided by an embodiment of the present application;
FIG. 2 is a schematic diagram showing the relationship between cis-position and click rate according to an embodiment of the present application;
FIG. 3 is a second flowchart of a media content recommendation method according to an embodiment of the present disclosure;
FIG. 4 is a block diagram of a media content recommendation device provided in an embodiment of the present application;
fig. 5 is a block diagram of an electronic device according to an embodiment of the present application.
Detailed Description
The following description of the embodiments of the present application will be made clearly and fully with reference to the accompanying drawings, in which it is evident that the embodiments described are some, but not all, of the embodiments of the present application. All other embodiments, which can be made by one of ordinary skill in the art without undue burden from the present disclosure, are within the scope of the present disclosure.
The terms "first," "second," and the like in this application are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus. Furthermore, the use of "and/or" in this application means at least one of the connected objects, such as a and/or B and/or C, is meant to encompass the 7 cases of a alone, B alone, C alone, and both a and B, both B and C, both a and C, and both A, B and C.
In the embodiments of the present application, words such as "exemplary" or "such as" are used to mean serving as examples, illustrations, or descriptions. Any embodiment or design described herein as "exemplary" or "for example" should not be construed as preferred or advantageous over other embodiments or designs. Rather, the use of words such as "exemplary" or "such as" is intended to present related concepts in a concrete fashion.
Referring to fig. 1, fig. 1 is a flowchart of a media content recommendation method according to an embodiment of the present application. The media content recommendation method may be executed by a server device, where the server device may be a cloud computer, a server, or other devices or data platforms with data processing functions, and the server executes the media content recommendation method as an example.
As shown in fig. 1, the media content recommendation method may include the steps of:
step 101, obtaining feedback information of a user on target media content.
The media content may also be called media resources (simply called media content), media data, etc., and may specifically include news, advertisement, video, audio, business, merchandise, etc., where the video may include video content such as a television show, a movie, a documentary, a variety, etc.
The target media content is a predetermined media content which is recommended to the client to increase the exposure, and the server can acquire feedback information of the user on the target media content at the client. The client may be a client on a mobile electronic device such as a mobile phone or a tablet, or may be a client on a large-screen electronic device such as a television or a projector, which may be specifically determined according to the actual situation, and the embodiments of the present application are not limited herein.
In particular implementations, the target media content may be predetermined based on current day hotspots or market demand. For example, if the national celebration festival is close, for celebration, the target media content may be video content of topics such as national celebration and republic; alternatively, if the actor a is going to be a table commemorative, the target media content may be video content corresponding to a representative movie work of the actor a.
The server may recommend the target media content to the client, which may be presented in the home page of the client, for example, or when keywords associated with the target media content are triggered, the target media content may be presented in the search results page of the client. The server may monitor the target media content, or alternatively, the server may monitor the target media content according to identification information, such as a name or a number, of the target media content, and collect feedback information of the user on the target media content at the client.
The feedback information may comprise positive feedback information of the user on the target media content, i.e. feedback information characterizing the user's interest in the target media content. In an example, the positive feedback information may include information that the user performs a click operation on the target media content, such as a number of clicks, or a frequency of clicks, a click rate, etc. over a period of time; information may also be included that the user performs collection, praise, coin-in, etc. operations on the target media content. The feedback information may also include negative feedback information of the user on the target media content, i.e. feedback information characterizing that the user is not interested in the target media content. In an example, the negative feedback information may include information that the user is not interested in performing, reducing recommendations, dislikes, etc., of the target media content. In addition, the feedback information may further include other feedback information of the target media content, such as comment, rating, bullet screen, and the like. It may be appreciated that the feedback information may include response information of any user to the target media content, which may be specifically determined according to practical situations, and embodiments of the present application are not limited herein.
The feedback information can reflect the interest degree of the target media content of the user, and the server can determine the user flow which can be brought by the target media content according to the feedback information. The server may then proceed to step 102.
Step 102, determining a correction coefficient of the target media content according to the feedback information and the preset feedback information when the feedback information is not matched with the preset feedback information, wherein the preset feedback information is associated with display information of the target media content at a client, and the display information comprises display position information and/or display time information.
In particular, the server may determine the preset feedback information according to the display information of the target media content currently on the client, that is, the expected user flow that may be brought by the display information of the target media content currently on the client. The display information may include display position information of the target media content at the client, for example, a display position of the page, a display order, or a size of a display area; the presentation information may also include presentation time information of the target media content at the client. The exposure degree provided by different display information is different, and the expected user flow brought by the different display information is different.
The server may compare the obtained feedback information with preset feedback information to determine whether the user traffic actually brought by the target media content meets the expectations. If the feedback information is not matched with the preset feedback information, the fact that the interested degree of the target media content of the user is not matched with the exposure degree provided by the current display information is indicated. Taking the display order of the target media content at the client as an example, if the feedback information of the target media content is lower than the expected feedback information, the user flow brought by the target media content is lower than the expected feedback information, and if the target media content is continuously placed at the higher display order, the exposure provided by the higher display order is wasted, and unnecessary network resources are occupied; if the feedback information of the target media content is higher than expected, the user flow rate brought by representing the target media content is higher than expected, and if the feedback information is continuously placed in a lower display order, the exposure of the target media content is suppressed, and a part of the user flow rate brought by the target media content is lost.
The server may correct the recommendation parameter of the target media content when the feedback information is not matched with the preset feedback information, where the recommendation parameter may include a recommendation score and/or a recommendation probability.
And step 103, correcting the recommended parameters of the target media content according to the correction coefficient.
In particular, the server may obtain the current recommendation score and/or recommendation probability of the target media content, for example, the recommendation score may determine, according to the correction coefficient and the first recommendation score of the target media content before correction, the second recommendation score of the target media content after correction. In an alternative embodiment, the server may determine the second recommendation score according to a product of the correction coefficient and the first recommendation score. The first recommendation score may be determined based on experience of an operator when the target media content is delivered for the first time, or in an alternative embodiment, the first recommendation score may be determined according to a recommendation score calculation method in the related art, for example, the recommendation score may be calculated according to an upper confidence algorithm (The Upper Confidence Bound Algorithm, UCB), specifically as follows:
wherein score i C for a first recommendation score for the target media content i,t T is the number of times the target media content is clicked in T experiments i,t The recommended number of times, or stated differently, the number of times the target media content was presented in t experiments. It is to be appreciated that the embodiment of determining the first recommendation score is not limited thereto, and the embodiments of the present application are not limited thereto.
After the corrected recommendation parameters are determined, the server can update the recommendation algorithm based on the corrected recommendation parameters, so that when the recommendation algorithm recommends the target media content, the server can perform recommendation based on the corrected recommendation parameters, and display information of the target media content on the client side is redetermined
In the embodiment of the application, after recommending the target media content to the client, the server can correct the recommendation parameters of the target media content by acquiring the feedback information of the user on the target media content, so that the target media content can be sufficiently exposed and promoted, has the recommendation parameters matched with the interesting degree of the user, and avoids occupying unnecessary recommendation resources, and improves the flexibility of media content recommendation.
In this embodiment of the present application, the target media content may be understood as media content that is recommended to the client among predetermined media content that needs to be increased in exposure or needs to be promoted.
Optionally, before the step 101, the method further includes:
determining a first media content;
recall the first media content based on a preset recall policy;
and determining the target media content in the first media content based on a preset screening strategy.
In this embodiment, the process of determining the target media content may include: the three stages of determining, recalling and screening are implemented, and in practical application, the screening stage can specifically comprise two links of sorting and screening. The following description will be given respectively:
1) The determining phase is for determining the first media content. The first media content is predetermined media content which needs to be increased in exposure or promoted, and can be regarded as an object which needs to be supported by the flow.
In an alternative embodiment, determining the first media content includes:
determining candidate media content;
determining estimated flow information of each candidate media content according to the characteristic information of each candidate media content;
and determining first media content in the candidate media content according to the estimated flow information of each candidate media content.
In particular, the selection range of the candidate media content may be predetermined. For example, if the national celebration festival is close, for the celebration, the selection range of the candidate media content may be the television resource and the movie resource of the romantic theme of the republic of 2000. After determining the candidate media content, traffic information for each of the candidate media content may be estimated, the traffic information being used to characterize the candidate media content's ability to attract users.
The feature information of the candidate media content may include: quality feature information, heat feature information, content feature information, production feature information, promotional feature information, and the like. The quality characteristic information is used for representing the quality condition of the candidate media content, optionally, the quality characteristic information comprises quality scores, and the server can determine the quality scores according to the playing time length, the scoring, the evaluation and other information of the candidate media content. The heat characteristic information is used for representing the heat condition of the candidate media content, optionally, the heat characteristic information comprises heat scores, and the server can determine the heat scores according to the information such as the playing time length, the clicking amount, the discussion degree, the number of viewers and the like of the candidate media content. The content characteristic information is used for representing the contents of the candidate media contents and can comprise characteristic information such as type contents, scenario contents, emotion contents and the like. The production feature information is used for representing the production situation of the candidate media content, and may include feature information of production personnel, such as feature information of actors, directors, drama and the like, and may also include feature information of production teams, such as feature information of production companies, feature information of post-companies and the like. The propaganda characteristic information is used for representing propaganda conditions of the candidate media content. It will be appreciated that the feature information of the candidate media content is not limited thereto, and may be specifically determined according to practical situations, and embodiments of the present application are not limited thereto.
The server may evaluate traffic information for each candidate media content based on the characteristic information.
In an alternative embodiment, the traffic information includes a traffic level, and the higher the traffic level of the candidate media content, the stronger the ability to attract users is characterized; the lower the traffic level of the candidate media content, the weaker the ability to characterize it to attract users. The server may determine the first candidate media content from among the candidate media content having the higher traffic class.
In another alternative embodiment, the traffic information includes traffic scores, and the server may input the feature information of each candidate media content into the traffic evaluation model through a pre-trained traffic evaluation model, and acquire the traffic scores output by the traffic evaluation model. The server may determine the first media content from among the candidate media contents whose traffic score satisfies the preset condition, for example, the server may select the candidate media content of the traffic score TopN to determine as the first media content.
2) The recall stage is used for determining a part of media content potentially interesting to the user from a mass media content library according to the characteristics of the user and the media content.
For a specific user to be recommended, since the first media content is a predetermined media content, if based on the recall policy in the related art, the first media content may not be all recalled. Thus, it is necessary to separately determine a recall policy, i.e. the preset recall policy, ensuring that the first media content is all recalled. It may be appreciated that in the embodiment of the present application, all the media contents recalled in the recall stage include a union of the first media content and the second media content, where the second media content is a recalled media content corresponding to the user to be recommended based on a recall policy of a related technology.
In an optional embodiment, the recalling the first media content based on a preset recall policy includes: and recalling the first media content according to the identification information of the first media content.
In this embodiment, after the first media content is determined, the identification information of the first media content may be stored separately, for example, in a redis system. In the recall stage, the server can singly open a recall path, acquire the identification information of the first media content in the redis system, and recall the first media content according to the identification information of the first media content.
3) The screening stage is used for screening the media content finally recommended to the user to be recommended after scoring and sorting the recalled media content.
The step of filtering the first media content to obtain the target media content and the step of filtering the second media content other than the first media content may be performed separately.
In specific implementation, the server may screen the first media content based on a preset screening policy to obtain the target media content, and determine an initial recommendation parameter of the target media content. The server may then recommend the target media content to the client based on the initial recommendation parameters.
In this embodiment of the present invention, after recommending the target media content to the client, the server may acquire feedback information of all users on the target media content, or may screen some users, and only acquire feedback information of the some users on the target media content.
Optionally, before the step 101, the method further includes:
determining a target user based on content tag information of target media content, wherein the user tag information of the target user is matched with the content tag information of the target media content;
The step 101 includes:
and acquiring feedback information of the target user on the target media content.
In this embodiment, the user of the feedback information acquired in step 101 is defined.
When the server monitors the target media content, only the feedback information of the target user on the target media content is obtained, and whether the feedback information is matched with preset feedback information or not is determined according to the feedback information of the target user.
The target user is a user who is matched with the target media content. On the premise of increasing the exposure of the target media content, the recommendation parameters of the target media content are corrected through the actual feedback condition of the user matched with the target media content, so that the data processing amount can be reduced, the capability of the target media content to actually attract the user can be determined from the perspective of the potential interested audience, the correction coefficient determined later is more accurate, and the accuracy of media content recommendation is further improved.
In a specific implementation, after determining the target media content, the server may acquire content tag information of the target media content, for example, video content, where the content tag information may include type tag information, theme tag information, emotion tag information, actor tag information, drama tag information, director tag information, producer tag information, and so on that characterize the target media content. The server may begin collecting user tag information after a new user registers the client, which may include gender tag information, age tag information, user preference tag information for video content, and the like.
The server may match the user according to the content tag information of the target media content, and if there is a match between at least one user tag information of a certain user and at least one content tag information of the target media content, the user may be determined to be the target user. Wherein the matching of the user tag information with the content tag information may include: the user tag information is the same as or similar to the content tag information, and may also include that the user tag information characterizes that the user is interested in the content corresponding to the content tag information, for example, the user marks the content tag information as "i prefer", "i favorite", "i look again later", and the like.
In the embodiment of the application, the server may continuously acquire feedback information of the user on the target media content, and perform information statistics. In order to further improve the flexibility of target media content recommendation, recommendation resources are utilized more fully, resource waste is avoided, the server can periodically count feedback information of a user on the target media content, and based on the feedback information of the target media content in a current counting period, a correction coefficient of the target media content in a next counting period is determined.
Optionally, the step 102 includes:
acquiring first feedback information of a user on the target media content in a first time period;
under the condition that the first feedback information is not matched with first preset feedback information, determining a first correction coefficient of the target media content according to the first feedback information and the first preset feedback information, wherein the first preset feedback information is preset feedback information corresponding to the first time period;
and correcting the recommended parameters of the target media content according to the correction coefficient, wherein the method comprises the following steps:
and correcting the recommended parameters of the target media content in a second time period according to the first correction coefficient, wherein the second time period is continuous with the first time period and is after the first time period.
The duration of the first time period and the second time period may be the same or different, and in an example, the first time period and the second time period may each be 1 hour.
In practical application, the user executes the clicking operation on the target media content, which indicates that the target media content attracts attention of the user, so that the clicking amount can reflect the response of the user to the target media content, and the capability of attracting the user of the target media content can be reflected.
In this embodiment of the present invention, optionally, the feedback information includes a click rate, the preset feedback information includes a preset click rate, and the server may determine whether the feedback information of the user on the target media content matches with the preset feedback information according to whether the actual click rate of the user on the target media content matches with the preset click rate, so as to determine whether the user traffic actually brought by the target media content meets the expectations.
In this embodiment, the server may determine the preset click rate according to the display information of the target media content currently on the client. Specifically, the preset click quantity is related to the display position information of the target media content at the client, taking the size of a display area as an example, the larger the display area is, the easier the user pays attention to the target media content, the larger the preset click quantity is, the smaller the display area is, the more difficult the user pays attention to the target media content, and the smaller the preset click quantity is; or taking the display order as an example, the more before the display order, the more focused the user to the target media content, the larger the preset click quantity, and the more after the display order, the more focused the user to the target media content, the smaller the preset click quantity.
Further alternatively, a graph of the relationship between the display information and the actual click rate may be drawn according to the statistical relationship between the multiple sets of display information and the actual click rate, and as shown in fig. 2, for example, the exposure and the click rate of the target media content will decrease with decreasing display order. And then, the server can obtain a functional relation between the display information and the preset click rate through fitting.
The preset click rate is also related to the display time information of the target media content at the client, and the click rate of the user on the target media content at 10 pm from 8 pm is generally higher than the click rate of the user on the target media content at 10 pm to 8 pm, so that the preset click rate corresponding to the peak time period is higher and the preset click rate corresponding to the low peak time period is lower. Further alternatively, the server may determine a corresponding preset click volume for each time period according to the time period.
In one example, the preset click rate is determined according to the following equation:
click j ＝exp(a·j b +c)
wherein click j And j is the display order of the target media content at the client for the preset click quantity, and the values of a, b and c are determined according to the display time of the target media content at the client. The server can determine a, b and c according to the characteristics of each time period, and correspondingly determine a function formula for calculating the preset click rate for each time period.
The server may compare the size of the click volume to the preset click volume to determine whether the click volume matches the preset click volume. Further optionally, the server may determine a preset click rate threshold based on the preset click rate, and determine that the click rate of the user on the target media content does not match the preset click rate when the click rate of the user on the target media content is less than or greater than the preset click rate threshold. Or, the server may determine a preset click rate range based on the preset click rate, and determine that the click rate of the target media content by the user does not match the preset click rate when the click rate of the target media content by the user is less than a lower limit value of the preset click rate range or greater than an upper limit value of the preset click rate range.
The preset click rate threshold may be equal to the preset click rate; or, the server may preset a threshold λ, and determine the preset click amount threshold according to the preset click amount and the threshold λ, where, for example, the preset click amount threshold is a sum of the preset click amount and the threshold λ, and the threshold λ may be positive or negative. For ease of understanding, it is assumed that the preset click rate is 12000, and the threshold λ may be-2000, which indicates that if the click rate of the user on the target media content is less than 10000, it is considered that the user traffic brought by the target media content does not reach the expected level; the threshold lambda may be 2000, which indicates that if the user clicks on the target media content is less than 14000, the user traffic caused by the target media content is not expected.
In an alternative embodiment, in the case that the click rate does not match the preset click rate, the correction coefficient of the target media content is determined according to the ratio of the click rate to the preset click rate.
In this embodiment, when the click rate is smaller than the preset click rate, the correction coefficient is smaller than 1, and the exposure that can be provided according to the recommended parameter corrected by the correction coefficient is smaller than the exposure that can be provided according to the recommended parameter before correction; and under the condition that the click rate is larger than the preset click rate, the correction coefficient is larger than 1, and the exposure provided by the recommended parameter corrected according to the correction coefficient is larger than the exposure provided by the recommended parameter before correction. In this way, when the actual click rate of the user on the target media content is higher than expected, the exposure of the target media content can be increased by correcting the recommendation parameter, and more user traffic is attracted, and when the actual click rate of the user on the target media content is lower than expected, the exposure of the target media content can be reduced by correcting the recommendation parameter, so that the recommendation parameter with higher exposure is reserved for the media content with higher actual click rate.
For ease of understanding, a complete example of embodiments of the present application is presented herein:
as shown in fig. 3, the specific implementation flow of this example is as follows:
step 301: and determining candidate media contents, performing flow estimation according to the characteristic information of the candidate media contents, and selecting the media contents of the flow rate Top20 as first media contents.
In this step, the server may extract the quality score, the popularity score, the actor, the director, the introduction, the poster information, the production company, and other feature information of each candidate media content, and estimate the flow of each candidate media content through a pre-trained flow estimation model. The quality score and the heat score can be obtained based on preset parameters and models, and the score can be obtained by obtaining relevant information on each forum website or social networking site.
According to the flow score output by the flow estimation model, the server may select the candidate media content of the flow score Top20 as the first media content.
Step 302: in a recall phase, the first media content is forced to be recalled.
In this step, the server may recall the first media content with a single-path policy in the recall stage based on the preset recall policy, and return the first media content together with other recall results based on the conventional recall policy. The server may then determine the target media content based on the preset screening policy.
Step 303: and screening the target user through the label information of the first media content.
In this step, the server may match the content tag information of the first media content with the user tags of all the users on the current website, and if the user tag information of a certain user matches with at least one content tag information of at least one first media content, determine the user as a target user.
Step 304: based on the target user, a score of the first media content is calculated according to a UCB algorithm in the exploration and utilization algorithm, and the target media content is determined.
In this step, the score of the first media content i i Can be expressed as:
wherein C is i,t T is the number of times the first media content was clicked in T experiments i,t The recommended number of times, or the number of times that the first media content is presented, in t experiments.
Sorting according to the score of each first media content, screening target media content based on sorting, and recommending the target media content to the client.
Step 305: and monitoring the target media content in real time.
In this step, the server monitors the click rate of the target media content in real time, and determines whether the real-time click rate matches with the preset click rate every hour, if not, a prompt signal is sent out, and step 306 is executed; if so, step 305 is repeated.
Specifically, the server may predict click volumes of different display orders in advance according to hours. According to data statistics, it can be determined that the higher the display order is, the higher the exposure and the click rate are, as shown in fig. 2, the server uses a mathematical formula to fit to obtain the relationship between the display order and the preset click rate, and the specific relationship is as follows:
click j ＝exp(a·j b +c)
wherein click j And j is the display order of the target media content at the client for the preset click quantity, and the values of a, b and c are related according to the display time period of the target media content at the client. Further, the server may determine the values of a, b, and c, respectively, according to the characteristics of each hour period. Then, in different hours, different relational expressions can be determined based on different values of a, b and c, and further different presets can be determinedClick volume.
The server may count the real-time click-through of the target media content i once an hour i 。
The server may click the real-time click of the target media content i during the current hour period i Click with preset click rate j In contrast, if click i ＜click j And +lambda, whether the real-time click quantity is matched with the preset click quantity or not is considered, the user flow actually brought by the target media content i is lower than expected, and a loss stopping signal is sent.
Step 306: and carrying out weight-reducing pressing on the target media content sending out the prompt signal.
In this step, the server may calculate the weight-reduction coefficient γ i ：
In the next hour period, a score click of the target media content is calculated i Then multiply by gamma i And obtaining the corrected score. In this way, the recommendation parameters for the target media content may be redetermined based on the revised score during the next hour period. Step 305 may then be repeated.
Referring to fig. 4, fig. 4 is a block diagram of a media content recommendation device according to an embodiment of the present application.
As shown in fig. 4, the media content recommendation apparatus 400 includes:
a first obtaining module 401, configured to obtain feedback information of a user on a target media content;
a first determining module 402, configured to determine, according to the feedback information and the preset feedback information, a correction coefficient of the target media content if the feedback information does not match the preset feedback information, where the preset feedback information is associated with display information of the target media content at a client, and the display information includes display position information and/or display time information;
and the correction module 403 is configured to correct the recommended parameters of the target media content according to the correction coefficient.
Optionally, the media content recommendation device 400 further includes:
the second determining module is used for determining a target user based on content tag information of target media content, wherein the user tag information of the target user is matched with the content tag information of the target media content;
the first acquisition module 401 is configured to:
and acquiring feedback information of the target user on the target media content.
Optionally, the first determining module 402 includes:
the first acquisition unit is used for acquiring first feedback information of the user on the target media content in a first time period;
the first determining unit is configured to determine, according to the first feedback information and the first preset feedback information, a first correction coefficient of the target media content when the first feedback information is not matched with the first preset feedback information, where the first preset feedback information is preset feedback information corresponding to the first time period;
the correction module 403 is configured to:
and correcting the recommended parameters of the target media content in a second time period according to the first correction coefficient, wherein the second time period is continuous with the first time period and is after the first time period.
Optionally, the feedback information includes a click rate, and the preset feedback information includes a preset click rate;
the first determining module 402 is configured to:
and under the condition that the click rate is not matched with the preset click rate, determining a correction coefficient of the target media content according to the ratio of the click rate to the preset click rate.
Optionally, the preset click rate is determined according to the following formula:
click j ＝exp(a·j b +c)
wherein click j For the pre-treatment ofSetting the click quantity, j as the display order of the target media content at the client, and determining the values of a, b and c according to the display time of the target media content at the client.
Optionally, the correction module 403 includes:
a second obtaining unit, configured to obtain a first recommendation score of the target media content before correction;
and the second determining unit is used for determining a second recommendation score of the target media content after correction according to the product of the correction coefficient and the first recommendation score.
Optionally, the media content recommendation device 400 further includes:
a third determining module for determining the first media content;
a recall module for recalling the first media content based on a preset recall policy;
and the screening module is used for determining the target media content in the first media content based on a preset screening strategy.
Optionally, the third determining module includes:
a fourth determining unit configured to determine candidate media contents;
a fifth determining unit, configured to determine estimated traffic information of each candidate media content according to feature information of each candidate media content;
and a sixth determining unit, configured to determine a first media content in the candidate media contents according to the estimated traffic information of each candidate media content.
Optionally, the recall module is configured to:
and recalling the first media content according to the identification information of the first media content.
The media content recommendation device 400 can implement the respective processes of the method embodiment corresponding to fig. 1, and achieve the same beneficial effects, and in order to avoid repetition, a detailed description is omitted here.
The embodiment of the application also provides electronic equipment. Referring to fig. 5, the electronic device 500 may include a processor 501, a memory 502, and a computer program 5021 stored in the memory 502 and capable of running on the processor 501, where the computer program 5021, when executed by the processor 501, can implement any steps in the above method embodiments and achieve the same beneficial effects, and will not be described herein.
Those of ordinary skill in the art will appreciate that all or a portion of the steps of implementing the methods of the embodiments described above may be implemented by hardware associated with program instructions, where the program may be stored on a readable medium. The embodiment of the application further provides a readable storage medium, on which a computer program is stored, where the computer program when executed by a processor can implement any step in the above method embodiment, and the same technical effects can be achieved, so that repetition is avoided, and no further description is given here.
Such as Read-Only Memory (ROM), random access Memory (Random Access Memory, RAM), magnetic or optical disk, etc.
While the foregoing is directed to the preferred embodiments of the present application, it will be appreciated by those of ordinary skill in the art that numerous modifications and variations can be made without departing from the principles set forth herein, and such modifications and variations are to be regarded as being within the scope of the present application.

Claims (20)

1. A method of media content recommendation, comprising:
acquiring feedback information of a user on target media content;
under the condition that the feedback information is not matched with preset feedback information, determining a correction coefficient of the target media content according to the feedback information and the preset feedback information, wherein the preset feedback information is associated with display information of the target media content at a client, and the display information comprises display position information and/or display time information;
and correcting the recommended parameters of the target media content according to the correction coefficient.
2. The method of claim 1, wherein prior to the obtaining user feedback information for the target media content, the method further comprises:
Determining a target user based on content tag information of target media content, wherein the user tag information of the target user is matched with the content tag information of the target media content;
the obtaining feedback information of the user on the target media content comprises the following steps:
and acquiring feedback information of the target user on the target media content.
3. The method of claim 1, wherein the determining the correction factor for the target media content based on the feedback information and the preset feedback information if the feedback information does not match the preset feedback information comprises:
acquiring first feedback information of a user on the target media content in a first time period;
under the condition that the first feedback information is not matched with first preset feedback information, determining a first correction coefficient of the target media content according to the first feedback information and the first preset feedback information, wherein the first preset feedback information is preset feedback information corresponding to the first time period;
and correcting the recommended parameters of the target media content according to the correction coefficient, wherein the method comprises the following steps:
and correcting the recommended parameters of the target media content in a second time period according to the first correction coefficient, wherein the second time period is continuous with the first time period and is after the first time period.
4. The method of claim 1, wherein the feedback information comprises a click volume and the preset feedback information comprises a preset click volume;
and under the condition that the feedback information is not matched with preset feedback information, determining the correction coefficient of the target media content according to the feedback information and the preset feedback information, including:
and under the condition that the click rate is not matched with the preset click rate, determining a correction coefficient of the target media content according to the ratio of the click rate to the preset click rate.
5. The method of claim 4, wherein the predetermined click rate is determined according to the following equation:
click j ＝exp(a·j b +c)
wherein click j And j is the display order of the target media content at the client for the preset click quantity, and the values of a, b and c are determined according to the display time of the target media content at the client.
6. The method of claim 1, wherein modifying the recommended parameters of the target media content based on the modification coefficients comprises:
acquiring a first recommendation score of the target media content before correction;
and determining a second recommendation score of the target media content after correction according to the product of the correction coefficient and the first recommendation score.
7. The method of claim 1, wherein prior to the obtaining user feedback information for the target media content, the method further comprises:
determining a first media content;
recall the first media content based on a preset recall policy;
and determining the target media content in the first media content based on a preset screening strategy.
8. The method of claim 7, wherein determining the first media content comprises:
determining candidate media content;
determining estimated flow information of each candidate media content according to the characteristic information of each candidate media content;
and determining first media content in the candidate media content according to the estimated flow information of each candidate media content.
9. The method of claim 7 or 8, wherein the recalling the first media content based on a preset recall policy comprises:
and recalling the first media content according to the identification information of the first media content.
10. A media content recommendation device, comprising:
the first acquisition module is used for acquiring feedback information of a user on target media content;
The first determining module is used for determining a correction coefficient of the target media content according to the feedback information and the preset feedback information when the feedback information is not matched with the preset feedback information, wherein the preset feedback information is associated with display information of the target media content at a client, and the display information comprises display position information and/or display time information;
and the correction module is used for correcting the recommended parameters of the target media content according to the correction coefficient.
11. The apparatus of claim 10, wherein the apparatus further comprises:
the second determining module is used for determining a target user based on content tag information of target media content, wherein the user tag information of the target user is matched with the content tag information of the target media content;
the first acquisition module is used for:
and acquiring feedback information of the target user on the target media content.
12. The apparatus of claim 10, wherein the first determining module comprises:
the first acquisition unit is used for acquiring first feedback information of the user on the target media content in a first time period;
The first determining unit is configured to determine, according to the first feedback information and the first preset feedback information, a first correction coefficient of the target media content when the first feedback information is not matched with the first preset feedback information, where the first preset feedback information is preset feedback information corresponding to the first time period;
the correction module is used for:
and correcting the recommended parameters of the target media content in a second time period according to the first correction coefficient, wherein the second time period is continuous with the first time period and is after the first time period.
13. The apparatus of claim 10, wherein the feedback information comprises a click volume and the preset feedback information comprises a preset click volume;
the first determining module is used for:
and under the condition that the click rate is not matched with the preset click rate, determining a correction coefficient of the target media content according to the ratio of the click rate to the preset click rate.
14. The apparatus of claim 13, wherein the predetermined click rate is determined according to the following equation:
click j ＝exp(a·j b +c)
wherein click j And j is the display order of the target media content at the client for the preset click quantity, and the values of a, b and c are determined according to the display time of the target media content at the client.
15. The apparatus of claim 10, wherein the correction module comprises:
a second obtaining unit, configured to obtain a first recommendation score of the target media content before correction;
and the second determining unit is used for determining a second recommendation score of the target media content after correction according to the product of the correction coefficient and the first recommendation score.
16. The apparatus of claim 10, wherein the apparatus further comprises:
a third determining module for determining the first media content;
a recall module for recalling the first media content based on a preset recall policy;
and the screening module is used for determining the target media content in the first media content based on a preset screening strategy.
17. The apparatus of claim 16, wherein determining a third determination module comprises:
a fourth determining unit configured to determine candidate media contents;
a fifth determining unit, configured to determine estimated traffic information of each candidate media content according to feature information of each candidate media content;
and a sixth determining unit, configured to determine a first media content in the candidate media contents according to the estimated traffic information of each candidate media content.
18. The apparatus of claim 16 or 17, wherein the recall module is to:
and recalling the first media content according to the identification information of the first media content.
19. An electronic device comprising a processor, a memory and a computer program stored on the memory and executable on the processor, which when executed by the processor performs the steps of the method according to any one of claims 1 to 9.
20. A computer-readable storage medium, characterized in that it has stored thereon a computer program which, when executed by a processor, implements the steps of the method according to any of claims 1 to 9.
CN202210066809.9A 2022-01-20 2022-01-20 Media content recommendation method and device, electronic equipment and readable storage medium Pending CN116521974A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210066809.9A CN116521974A (en) 2022-01-20 2022-01-20 Media content recommendation method and device, electronic equipment and readable storage medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210066809.9A CN116521974A (en) 2022-01-20 2022-01-20 Media content recommendation method and device, electronic equipment and readable storage medium

Publications (1)

Publication Number Publication Date
CN116521974A true CN116521974A (en) 2023-08-01

Family

ID=87396332

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210066809.9A Pending CN116521974A (en) 2022-01-20 2022-01-20 Media content recommendation method and device, electronic equipment and readable storage medium

Country Status (1)

Country Link
CN (1) CN116521974A (en)

Similar Documents

Publication Publication Date Title
US8151194B1 (en) Visual presentation of video usage statistics
US9715731B2 (en) Selecting a high valence representative image
US9235574B2 (en) Systems and methods for providing media recommendations
US11620678B2 (en) Advertising method, device and system, and computer-readable storage medium
CN103686237B (en) Recommend the method and system of video resource
US8898696B2 (en) System and method for allocating advertisements
US10380649B2 (en) System and method for logistic matrix factorization of implicit feedback data, and application to media environments
US20150332373A1 (en) Method and system for pushing mobile application
CN111178970B (en) Advertisement putting method and device, electronic equipment and computer readable storage medium
CN109753601B (en) Method and device for determining click rate of recommended information and electronic equipment
KR20160055930A (en) Systems and methods for actively composing content for use in continuous social communication
CN102523511A (en) Network program aggregation and recommendation system and network program aggregation and recommendation method
CN106131703A (en) A kind of method of video recommendations and terminal
JP2008542870A (en) Method and apparatus for estimating the overall interest of a group of users for content
CN109168047B (en) Video recommendation method and device, server and storage medium
US8559749B2 (en) Audiovisual content delivery system
CN105338408B (en) Video recommendation method based on time factor
CN113656681B (en) Object evaluation method, device, equipment and storage medium
CN108021700A (en) A kind of user tag generation method, device and server
Xu et al. Catch-up TV recommendations: show old favourites and find new ones
US20190050890A1 (en) Video dotting placement analysis system, analysis method and storage medium
KR102111491B1 (en) System and Method for Estimating the Indirect Advertising Value of Independent Media Video Channel
CN109688217B (en) Message pushing method and device and electronic equipment
CN109213933B (en) Content item recommendation method, device, equipment and storage medium
CN113204699B (en) Information recommendation method and device, electronic equipment and storage medium

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination