WO2018090793A1 - Multimedia recommendation method and device - Google Patents

Abstract

A multimedia recommendation method and a device, in the method, after determining multimedia resources to be recommended or to be released, obtain a credit score of a user related to the multimedia resources to be recommended or to be released; determine, according to the credit score of the user related to the multimedia resources to be recommended or to be released, a sorting order of multimedia resources to be recommended or to be released; and recommend the multimedia resources to a terminal on the basis of the sorting order of the multimedia resources to be recommended or to be released. This method can improve click-rate of the multimedia resources and promote the rapid spread of the multimedia resources.

Description

Multimedia recommendation method and device

The present application claims priority to Chinese Patent Application No. 201611019026.6, entitled "A Multimedia Recommendation Method and Apparatus" on November 18, 2016, the entire contents of which are incorporated herein by reference. .

Technical field

The present application relates to the field of network communication technologies, and in particular, to multimedia recommendation.

Background technique

With the continuous development of network technology, the number of multimedia resources that network platforms can provide has gradually increased. Users can use mobile phones, computers and other terminals to access the network platform to view the multimedia resources existing on the network platform.

With the continuous development of the society, users are increasingly demanding high-quality multimedia resources such as hot and interesting. However, the diversity of multimedia resources in the network platform is large, and users may need a large amount of search. Locating high-quality multimedia resources takes a long time. At the same time, if the multimedia resources displayed for users do not meet the user's viewing requirements, it will also affect the propagation of multimedia resources, resulting in the waste of multimedia resources. Therefore, how to more reasonablely recommend multimedia resources for users and promote the dissemination of multimedia resources is a technical problem that those skilled in the art urgently need to solve.

Summary of the invention

In view of this, the embodiment of the present application provides a multimedia recommendation method and apparatus, so as to improve the click rate of multimedia resources and promote the propagation of multimedia resources.

To achieve the above objective, the first aspect of the embodiments of the present application provides a multimedia recommendation method, including:

Receiving a page presentation request sent by the terminal, where the page presentation request is for requesting the multimedia presentation page;

Determining at least one multimedia resource to be recommended;

Determining the quality score of the multimedia resource to be recommended, according to the credit score of the historical access user of the multimedia resource to be recommended, the number of historical access users of the multimedia resource to be recommended, and the preset scoring rule, , the default scoring rule is the calendar of multimedia resources The number of visited users is positively correlated with the quality score of the multimedia resource, and/or the credit score of the historical access user of the multimedia resource is positively correlated with the quality score of the multimedia resource;

Sorting the recommended multimedia resources according to the highest to lowest quality scores of the multimedia resources to be recommended;

The multimedia presentation page is fed back to the terminal, and the multimedia resource to be recommended is recommended according to the order of the multimedia resources to be recommended in the multimedia presentation page.

A second aspect of the embodiments of the present application provides a multimedia recommendation method, including:

Obtaining at least one multimedia resource to be released, and an identifier of a publisher of each of the to-be-published multimedia resources;

For each of the to-be-published multimedia resources, determining a credit score of the publisher according to the identifier of the publisher;

Sorting the to-be-published multimedia resources according to a ranking of the publisher's credit scores of the to-be-published multimedia resources from high to low;

Determining, by the sorting order of the to-be-published multimedia resources, a recommendation order of recommending the to-be-published multimedia resources to the terminal.

A third aspect of the embodiments of the present application provides a multimedia recommendation method, including:

The server receives a page presentation request sent by the terminal, and the page presentation request is used to request the multimedia presentation page;

The server determines at least one multimedia resource to be recommended;

For each multimedia resource to be recommended, the server determines the quality score of the multimedia resource to be recommended according to the credit score of the historical access user of the to-be-recommended multimedia resource, the number of historical access users of the multimedia resource to be recommended, and a preset scoring rule. The preset scoring rule is that the number of historical access users of the multimedia resource is positively correlated with the quality score of the multimedia resource, and/or the credit score of the historical access user of the multimedia resource is positively correlated with the quality score of the multimedia resource;

The server sorts the recommended multimedia resources according to the highest to lowest order of the quality scores of the multimedia resources to be recommended;

The server feeds back the multimedia presentation page to the terminal, and recommends the multimedia resource to be recommended according to the order of the multimedia resources to be recommended in the multimedia presentation page.

A fourth aspect of the embodiments of the present application provides a multimedia recommendation method, including:

The server acquires at least one multimedia resource to be released, and an identifier of the publisher of each multimedia resource to be released;

The server determines, for each multimedia resource to be released, a credit score of the publisher according to the identifier of the publisher;

The server sorts the multimedia resources to be distributed according to the highest priority of the publisher's credit score of the multimedia resource to be released;

The server determines the sorting order of the multimedia resources to be released, and determines the recommended order of the multimedia resources to be released to the terminal.

A fifth aspect of the embodiments of the present application provides a multimedia recommendation apparatus, including:

a request receiving unit, configured to receive a page presentation request sent by the terminal, where the page presentation request is used to request the multimedia presentation page; and determine at least one multimedia resource to be recommended;

a quality determining unit, configured to determine, according to the credit score of the historical access user of the multimedia resource to be recommended, the number of historical access users of the multimedia resource to be recommended, and a preset scoring rule, for each multimedia resource to be recommended, to determine the multimedia to be recommended The quality score of the resource, wherein the preset scoring rule is that the number of historical access users of the multimedia resource is positively correlated with the quality score of the multimedia resource, and/or the credit score of the historical access user of the multimedia resource is positively correlated with the quality score of the multimedia resource. ;

a resource sorting unit, configured to sort the recommended multimedia resources according to a highest to lowest quality score of the multimedia resources to be recommended;

The resource presentation unit is configured to feed back the multimedia presentation page to the terminal, and recommend the multimedia resource to be recommended according to the order of the multimedia resources to be recommended in the multimedia presentation page.

A sixth aspect of the embodiments of the present application provides another multimedia recommendation apparatus, including:

An information acquiring unit, configured to acquire at least one multimedia resource to be released, and an identifier of a publisher of each multimedia resource to be released;

a credit determining unit, configured to determine, according to the identifier of the publisher, a credit score of the publisher for each multimedia resource to be released;

a resource sorting unit, configured to sort the multimedia resources to be distributed according to the highest to lowest order of the credit scores of the publishers of the multimedia resources to be released;

a recommendation processing unit, configured to determine a sort order of the multimedia resource to be released, to recommend to the terminal The recommended order of the multimedia resources to be released.

A seventh aspect of the embodiments of the present application provides a server, including:

Processor and memory;

a memory for storing program code and transmitting the program code to the processor;

The processor is configured to invoke the instruction in the memory to perform the multimedia recommendation method provided by the first aspect or the third aspect of the embodiment of the present application.

An eighth aspect of the embodiment of the present application provides another server, including:

Processor and memory;

a memory for storing program code and transmitting the program code to the processor;

The processor is configured to invoke the instruction in the memory to perform the multimedia recommendation method provided by the second aspect or the fourth aspect of the embodiment of the present application.

A ninth aspect of the present application provides a storage medium, where the storage medium is used to store a program code, and the program code is used to execute the multimedia recommendation method provided by the first aspect or the third aspect of the embodiment of the present application.

A tenth aspect of the embodiment of the present application provides another storage medium, where the storage medium is used to store a program code, and the program code is used to execute the multimedia recommendation method provided by the second aspect or the fourth aspect of the embodiment of the present application.

The eleventh aspect of the present application provides a computer program product, including instructions, which, when run on a server, causes the server to perform the multimedia recommendation method provided by the first aspect or the third aspect of the embodiment of the present application.

A twelfth aspect of the present application provides another computer program product, including instructions, which, when run on a server, causes the server to perform the multimedia recommendation method provided by the second aspect or the fourth aspect of the embodiment of the present application.

It can be seen from the above that in the embodiment of the present application, after determining the multimedia resource to be recommended or to be released, the server may obtain the credit score of the user related to the multimedia resource, and treat the credit score of the user related to the multimedia resource according to the credit score of the user related to the multimedia resource. The multimedia resources are recommended or to be sorted, and the credit score of the user related to the multimedia resource may reflect the quality of the multimedia resource, so that the multimedia resources of the user related to the multimedia resource may sort the plurality of multimedia resources, and the quality may be compared. High-order multimedia resources are ranked higher, so that the top-ranked multimedia resources can be pushed first. It is recommended to the terminal, so that the user of the terminal can preferentially see the high-quality multimedia resources, so that it does not need to spend a lot of time to search for the multimedia resources, which is convenient for the user to access the multimedia resources quickly and conveniently, and also improves the user's click on the multimedia resources. The rate has promoted the spread of multimedia resources.

DRAWINGS

In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings to be used in the embodiments or the prior art description will be briefly described below. Obviously, the drawings in the following description are only It is an embodiment of the present application, and those skilled in the art can obtain other drawings according to the provided drawings without any creative work.

FIG. 1 is a schematic structural diagram of a possible structure of a multimedia recommendation system according to an embodiment of the present application;

2 is a schematic flowchart of an embodiment of a multimedia recommendation method according to an application of the present application;

3 is a schematic flowchart of still another embodiment of a multimedia recommendation method according to the present application;

4 is a schematic diagram of a training process of a training credit scoring model according to the present application;

FIG. 5 is a schematic flowchart diagram of another multimedia recommendation method according to the present application;

6 is a schematic structural diagram of an embodiment of a multimedia recommendation apparatus according to the present application;

FIG. 7 is a schematic structural diagram of an embodiment of another multimedia recommendation apparatus of the present application; FIG.

FIG. 8 is a schematic structural diagram of a server embodiment of the present application; FIG.

FIG. 9 is a schematic diagram showing the structure of another server embodiment of the present application.

detailed description

The technical solutions in the embodiments of the present application are clearly and completely described in the following with reference to the drawings in the embodiments of the present application. It is obvious that the described embodiments are only a part of the embodiments of the present application, and not all of the embodiments. All other embodiments obtained by a person of ordinary skill in the art based on the embodiments of the present application without departing from the inventive scope are the scope of the present application.

The multimedia recommendation method in the embodiment of the present application can be applied to the recommendation of multimedia resources such as audio and video.

FIG. 1 is a schematic diagram showing the structure of a multimedia recommendation system of the present application. The system may include: a media service platform 10 and at least one terminal 11.

The media service platform may include at least one server 101 and at least one database 102, and the server 101 may be connected to the database 102 through a network.

In order to improve the processing efficiency of the multimedia presentation request and the multimedia recommendation, the media service platform may include a server cluster composed of a plurality of servers 101.

In the embodiment of the present application, the server 101 may be configured to process a multimedia presentation request of a client in the terminal, such as processing a page presentation request for requesting a presentation page including multimedia; processing a playback request of the multimedia resource initiated by the terminal, and the terminal is to the media. The service platform uploads multimedia resources or requests to download multimedia resources, and the like.

In the embodiment of the present application, the server 101 is further configured to collect behavior data of the terminal user accessing the media service platform, and store the collected behavior data as the historical behavior data of the terminal user in a designated storage area of the server, or Historical behavior data is stored in the database.

The database 102 can store historical behavior data of each user in the media service platform collected by the server, and can also store multimedia resources in the media service platform and the like.

In the embodiment of the present application, the terminal 11 can be equipped with a client where the player is located; or a client where the browser is located; or a client of a server in another media service platform, for example, the client can be a The client of the media service platform uploading multimedia resources. Specifically, the terminal can be a mobile phone, a tablet computer, a desktop computer, or the like.

In the application embodiment, the terminal 11 may be configured to send a page presentation request to the server, where the page presentation request is used to request the multimedia presentation page to present a plurality of multimedia resources that are selectable for the user to select to play in the multimedia presentation page. For example, a recommendation list of multimedia may be included in the multimedia presentation interface, so that the user of the terminal selects the multimedia resource to be played from the recommendation list.

Correspondingly, the server 101 may be configured to receive a page presentation request sent by the terminal, where the page presentation request is used to request the multimedia presentation page, and determine at least one multimedia resource to be recommended; and for each multimedia resource to be recommended, according to the multimedia resource to be recommended. The historical access user's credit score, the number of historical access users of the to-be-recommended multimedia resource, and the preset scoring rule determine the quality score of the to-be-recommended multimedia resource, in descending order of the quality score of the multimedia resource to be recommended. The recommended multimedia resources are sorted; the multimedia presentation page is fed back to the terminal, and the multimedia resources to be recommended are recommended according to the order of the multimedia resources to be recommended in the multimedia presentation page. The preset scoring rule is that the number of historical access users of the multimedia resource and the quality score of the multimedia resource are positive. Relevant, and/or, the historical access user's credit score of the multimedia resource is positively correlated with the quality score of the multimedia resource.

After receiving the page presentation request sent by the terminal, the server obtains the quality score of the multimedia resource. The more the number of historical access users accessing the multimedia resource, the higher the credit score of the historical access user, and the quality score of the multimedia resource. The higher, in this way, according to the order of the quality scores of the multimedia resources from high to low, the ranking of the multimedia resources suitable for the mass and the higher quality can be made higher, so after feeding the multimedia presentation page to the terminal, In the multimedia presentation page, the multimedia resources are recommended according to the sorting order of the multiple multimedia resources, so that the high-quality multimedia resources are preferentially recommended, which is beneficial for the user of the terminal to quickly find high-quality multimedia resources, thereby facilitating the user to quickly and conveniently. Accessing multimedia resources also increases the enthusiasm of users to access multimedia resources, thereby increasing the click-through rate of multimedia resources and facilitating the rapid dissemination of multimedia resources.

The media server may generate a recommendation list according to a sorting order corresponding to each multimedia resource, wherein the higher the order of the multimedia resources in the recommendation list is, the higher the order is.

With reference to the above commonality, referring to FIG. 2, a schematic flowchart of an embodiment of a multimedia recommendation method is shown. The method in this embodiment may include:

S201. The terminal sends a page presentation request to a server in the media service platform.

The page presentation request is used to request a multimedia presentation page.

For example, taking the client where the terminal is equipped with the player as an example, when the user opens the player, the terminal sends a page presentation request to the server corresponding to the player, and also requests to display a presentation page containing multiple video resources.

For example, when the terminal is equipped with the client where the browser is located, when the user wants to display multimedia resources such as playable video through the browser, the terminal where the browser is located requests the multimedia presentation page from the server.

S202. The server determines at least one multimedia resource to be recommended.

The server can use all the multimedia resources in the media service platform as the multimedia resources to be recommended, and the multimedia resources to be recommended can be understood as the multimedia resources that can be recommended to the terminal. The server may also determine part of the multimedia resource as the multimedia resource to be recommended according to actual needs. For example, the server may determine the multimedia resource to be recommended according to the geographical location where the terminal is located, the access right of the terminal, and the like.

S203. For each multimedia resource to be recommended, the server determines to access a historical user set of the multimedia resource to be recommended.

The historical access user set includes at least one historical access user of the multimedia resource to be recommended. The historical access user refers to the user who has visited the multimedia resource to be recommended before receiving the page presentation request.

The historical access user accessing the multimedia resource to be recommended may reach the preset duration for accessing the multimedia resource to be recommended, in order to accurately determine the quality score of the multimedia resource to be recommended according to the history of the multimedia resource to be recommended. The preset duration can be determined according to actual needs and the maximum play duration corresponding to the multimedia resource to be recommended. For example, the video to be recommended is a video. The historical access user of a video is the user who has watched the video for a preset duration.

Further, the historical access user of the to-be-recommended multimedia resource may be a user who browses the multimedia resource completely. For example, the video is still taken as an example, and the historical user of the video is a user who completely views the video.

S204. The server obtains a credit score of each historical access user in the historical user set.

The credit score of the historical visiting user may be the credit score of the historical visiting user in the media service platform.

The historical access user's credit score may be determined based on historical performance data of the historical access user in the media service platform. The credit score of the historical visiting user is based on the historical behavior data of the historical visiting user, and the ranking relationship is determined by using the mapping relationship between the historical behavior data obtained by the pre-training and the credit score. Specifically, the credit score of the historical visiting user may be calculated based on the historical behavior data and using a credit scoring model reflecting the mapping relationship between the historical behavior data and the credit score.

A training process of a credit scoring model reflecting a mapping relationship between historical behavior data and credit scores may be: first, determining historical behavior data of a plurality of sample users in the media service platform and third party credit rating of the plurality of sample users Then, the third-party credit score of the plurality of sample users is used as the credit score of the sample user, and the regression analysis is performed according to the third-party credit score of the plurality of sample users and the historical behavior data of the plurality of sample users to determine A credit scoring model that reflects the mapping between historical behavioral data and credit scores.

Of course, the credit score of the historical visiting user may also be based on the user in the media service platform. The historical behavior data and the third party credit data are jointly determined.

It should be noted that the historical access user's credit score may be pre-calculated and stored in the server or the database; or the server may obtain the historical behavior data of each historical access user in turn after determining the historical user set, and access according to the history. The user's historical behavior data and the credit scoring model calculate the credit score of the historical visiting user in real time.

S205. Determine, for each of the to-be-recommended multimedia resources, a quality score of the to-be-recommended multimedia resource according to the credit score of the historical access user of the to-be-recommended multimedia resource, the number of historical access users of the to-be-recommended multimedia resource, and a preset scoring rule. .

Among them, the quality scoring model can be pre-built according to the scoring rules.

The preset scoring rule may positively correlate the number of historical access users of the multimedia resource with the quality score of the multimedia resource, and/or the credit score of the historical access user of the multimedia resource is positively correlated with the quality score of the multimedia resource.

For example, the preset scoring rule may be the number of historical access users for the multimedia resource, and the higher the credit score of the historical visiting user, the higher the quality score of the multimedia resource.

In order to satisfy the above-mentioned preset scoring rules, the first weight corresponding to the number of historical access users and the second weight corresponding to the credit score of the historical visiting user may be respectively set in the quality scoring model, and the historical visiting user is comprehensively determined. Quality score.

Of course, in addition to setting the number of historical access users and the weight of the credit score, the quality scoring model may also have the possibility of determining the quality score of the multimedia resource, such as the quality scoring model may calculate all historical visits corresponding to the multimedia resource. The sum of the user's credit scores, and the sum of the credit scores is determined as the quality score of the multimedia resource.

In this embodiment, for a multimedia resource to be recommended, the number of historical access users of the multimedia resource to be recommended and the credit score of each historical access user are used to determine the quality score of the multimedia resource to be recommended as an example. .

However, it can be understood that in the actual application, the preset scoring rule may have other possibilities. For example, the preset scoring rule may also be: the higher the credit score of the historical access user of the multimedia resource, the higher the quality score of the multimedia resource. In this case, the quality score of the multimedia resource may not be related to the number of historical access users of the multimedia resource. For example, the quality score of the multimedia is an average of the credit scores of all historical visiting users corresponding to the multimedia resource. Default rating rule It can also be: the more the number of historical access users of the multimedia resource, the higher the quality score of the multimedia resource. It should be noted that, in the embodiment of FIG. 2, after step S203 to step S205 are to determine at least one multimedia resource to be recommended, for each multimedia resource to be recommended, a historical access user accessing the multimedia resource to be recommended is determined in real time. The credit score of the multimedia resource to be recommended is determined in real time according to the number of historical access users of the multimedia resource to be recommended and the credit score of the historical access user of the multimedia resource to be recommended.

However, it can be understood that, in practical applications, the quality score of each multimedia resource to be recommended may also be pre-calculated and stored in a server or a database, in which case the server may directly collect from the stored quality score. The quality score corresponding to each multimedia resource to be recommended is queried. For each of the to-be-recommended multimedia resources, the quality score of the to-be-recommended multimedia resource may be calculated according to the number of historical access users accessing the to-be-recommended multimedia resource and the credit of the historical access user of each of the to-be-recommended multimedia resources. The score is determined by a quality scoring model that conforms to the preset scoring rules and will not be described here.

S206. The server sorts the recommended multimedia resources according to the highest to lowest quality score of the multimedia resources to be recommended.

The ordering of the plurality of multimedia resources to be recommended is to obtain a recommendation sequence for recommending the multimedia resources to be recommended.

S207. The server feeds back the multimedia presentation page to the terminal, and recommends the multimedia resource to be recommended according to the sort order of the multimedia resources to be recommended in the multimedia presentation page.

After the ranking of the step S206, the higher the quality score of the multimedia resource to be recommended, the higher the ranking of the multimedia resources to be recommended, and the order of the multimedia resources to be recommended is actually the recommended order of the multimedia resources to be recommended. In the multimedia presentation page that the server feeds back to the terminal, the more advanced multimedia resources to be recommended are preferentially recommended, so that the probability of being selected by the user is higher.

In practical applications, in addition to the quality score of multimedia resources affecting the user's demand for the multimedia resource, there are differences in the interests of different users, and the interests of different users may be through the user in the media service platform. The credit score is determined, and the study finds that the gaps in the hobbies of users with different scores in credit scores are larger, while the credit scores are smaller in the hobbies of users who want to pass the scores. Therefore, in order to be more reasonable Video recommendations to users, The recommendation of the video resource can be performed for the credit score based on different users and the quality score of the multimedia resource. For example, referring to FIG. 3, which is a schematic flowchart of still another embodiment of a multimedia recommendation method according to the present application, the method in this embodiment may include:

S301. The terminal sends a page presentation request to a server in the media service platform.

The page presentation request is used to request a multimedia presentation page.

The page shows that the request carries the identity of the target user who logs in to the server through the terminal. For example, the identifier of the target user may be an account number, a user name, and the like of the target user of the login server.

S302. The server obtains a credit score of the target user according to the identifier of the target user.

The process of obtaining the credit score of the target user may be similar to the process of obtaining the credit score of the historical access user in the embodiment of FIG. 2, for example, the credit score of the target user may be queried from the pre-stored credit score; The historical score data of the target user is queried in real time, and the credit score model reflecting the mapping relationship between the historical behavior data and the credit score is used to determine the credit score of the target user. For details, refer to related descriptions of the previous embodiments, and details are not described herein again.

S303. The server determines at least one multimedia resource to be recommended.

S304. The server determines a quality score of the to-be-recommended multimedia resource for each multimedia resource to be recommended.

For the steps S303 and S304, reference may be made to related descriptions of the foregoing embodiments, and details are not described herein again.

S305. For each multimedia resource to be recommended, the server determines a difference between the credit score of the target user and the quality score of the multimedia resource to be recommended.

S306. The server sorts the recommended multimedia resources according to the absolute value of the difference.

It can be understood that the user's credit score can reflect the user's interest preference, and the quality score of the multimedia resource is also determined according to the credit score of the historical visiting user who has accessed the multimedia resource, so that the quality score of the multimedia resource can not only be Reflecting the quality of the multimedia resources, it can also reflect the interest status of the group that likes the multimedia resource, so that the difference between the target user's credit score and the multimedia resource reflects the target user's preference for the multimedia resource. .

The difference between the credit score of the target user and the quality score of the multimedia resource may be negative, and if the quality score of the two multimedia resources is the absolute value of the difference between the credit score of the target user and the credit score of the target user The same is true, the degree of interest of the two multimedia resources is the same. Therefore, in this step S306, the plurality of multimedia resources to be recommended need to be sorted according to the absolute value of the difference.

The smaller the absolute value of the difference between the quality score of the multimedia resource and the credit score of the target user, the higher the degree of interest of the target user to the multimedia resource.

S307, in the process of sorting the recommended multimedia resources, the server uses at least two multimedia resources to be recommended with the same absolute value as the sorting set, and the quality scores of the multimedia resources to be recommended in the sorting set are in descending order. And sorting the multimedia resources to be recommended in the sorting set.

If the absolute values of the differences corresponding to the two or more multimedia resources are the same, in order to be able to recommend high-quality multimedia resources to the target user, the quality scores of the multimedia resources may be ranked in descending order of the two or The plurality of multimedia resources are sorted to provide high-quality multimedia resources for the target users on the premise of meeting the interests of the target users.

S308. The server feeds back the multimedia presentation page to the terminal, and recommends the multimedia resource to be recommended according to the order of the multimedia resources to be recommended in the multimedia presentation page.

In the above embodiments of the present application, the historical behavior data of the user may be data generated by the user in the process of historically accessing the media service platform. The historical access behavior data may include: a multimedia resource accessed by the user, a resource type and a content category of the multimedia resource accessed by the user, a duration of accessing the multimedia resource, and the like.

The user's historical behavior data may be historical behavior data of the user accessing multimedia resources of different media categories.

For example, according to the distribution channels and the content of multimedia resources, the media categories of multimedia resources can be divided into: variety, animation, politics, music, hilarity, documentary, online movie, sports, fashion, technology, finance, live broadcast, education, Real estate, maternal and child, travel, action movies, adventure movies, love movies, and many other media categories. In this way, for each media category, historical behavior data of the user accessing the multimedia resources under the media category can be separately stored.

Correspondingly, in the embodiment of the present application, the behavior data of the multimedia resource accessed by the user in a certain media category may also be multiple, for example, the number of accesses of the multimedia resource for the media category (View Times), View Duration, Session Starts, Session Duration, Session Ends, Upload Clicks of Uploaded Multimedia Resources, Upload Frequency of Multimedia Resources (Upload Frequency) and many more. The duration of the access stay can be understood as the total duration of accessing the multimedia resource, including the tentative time and the duration of the unplayed multimedia resource; the start of the play can be understood as the time when the multimedia resource starts to be played; correspondingly, the play duration can be understood as playing. The total duration of the multimedia resource (excluding the tentative time); the end of playback can be understood as the end of playback time.

Through the above historical behavior data, the user's hobbies can be reflected.

Further, through research, it is found that the credit scores of the users evaluated by the credit information platform can also objectively reflect the interests and interests of users. In general, the interest groups of the user groups with different scores in the credit scores are very different. Large, and the hobbies of the user groups whose credit scores are in the same score segment are relatively similar. For example, the user group of Gao Zhengxin points obviously has a high click-through rate in the types of finance, politics, art, etc., and the user group with low credit scores has a high click-through rate for funny, segmental and other types of content. Therefore, the embodiment of the present application may predetermine the sample user, and obtain the credit score of the sample user in the third-party platform, and then based on the sample user's credit score and historical behavior data, the training reflects the historical user behavior data and the user is A credit scoring model that maps the relationship between credit scores in the media service platform.

For ease of understanding, an implementation manner of training a credit scoring model for calculating a user's credit score using the historical behavior data of the sample user and the third-party credit score of the sample user is described below. For example, referring to FIG. 4, a schematic flowchart of a training credit scoring model of the present application is shown. The method of the present application may be applied to a server, and may include:

S401: Select a sample user from the media service platform, and obtain a behavior observation value xi, j of the access behavior of each sample user in the multimedia resource of different media categories.

The behavior observation value can be understood as the data value of the historical behavior data of the user accessing the multimedia resource.

For example, the variable xi,j can be used to represent the behavioral observations of the user in a certain type of multimedia resource. Where i is the media category of the multimedia resource accessed by the user, i∈(1, 2, 3, ..., n), where n is the upper limit of the behavior type of the user's access behavior, for example, the user's access behavior includes the foregoing The number of visits (View Times), the duration of the visit (View Duration), the start of playback, and so on. For the media category, refer to the related introduction of the previous embodiment. j represents the type of behavior of the user, for example, J∈("ViewTime","ViewDuration","UploadFre","UploadClick",...).

S402. Construct a matrix X corresponding to the historical behavior data of the sample user by using the behavioral observation value x _i,j of the sample user for different media categories.

Among them, the matrix X can be expressed as:

Where m is the maximum value of the media category. For example, if the media category has 5 categories, then m is equal to 5.

S403, converting the matrix X into a matrix X' suitable for regression estimation.

Where X' = (x ₁₁ ... x _1m x ₂₁ ... x _2m x ₃₁ ... x _m1 ... x _mn ).

S404: Obtain a credit score Y of each sample user at a third-party credit institution.

Where Y = (y ₁ , y ₂ , ..., y _n ) ^T , where y _i represents the credit score of the user i at the third party credit bureau.

S405: Calculate a formula for calculating a credit score of a user in the media service platform by using a matrix X′ corresponding to each sample user and obtaining a mapping relationship β of X′ to Y and a mapping error vector ε by linear regression.

For example, the regression formula can be written as:

Y=X'β+ε (Formula 1);

Where β is a mapping weight matrix, which can be represented by the following matrix:

β = (β ₁₁ ... β _1m β ₂₁ ... β _2m β ₃₁ ... β _m1 ... β _mn );

和

In this way, the estimated amount of β and ε is generated using the sample data.

with

和

代入到回归公式，得到用户在媒体服务平台中信用评分

的计算公式：Estimate

with

Substituting into the regression formula to get the user's credit score in the media service platform

Calculation formula:

After formula 2 is obtained, if it is necessary to calculate a credit score of a user in the media service platform, the historical behavior data of the user may be acquired first, and then a matrix X representing the historical behavior data of the user is generated, and the matrix X is input. Go to Equation 2 to get the user's credit score.

On the other hand, on the basis of the multimedia recommendation system shown in FIG. 1 , the embodiment of the present application further provides another multimedia recommendation method, where the multimedia recommendation method is suitable for recommending the latest release to the terminal. media resources.

For example, referring to FIG. 5, which is a schematic flowchart of another embodiment of a multimedia recommendation method according to the present application, the method of this embodiment may be applied to a server. The method of this embodiment may include:

S501. Acquire at least one multimedia resource to be released, and an identifier of a publisher of each multimedia resource to be released.

After the terminal uploads the multimedia resources that need to be released to the server, the server needs to review the multimedia resources to be released and publish them after the audit. For example, the server may determine the multimedia resource to be released once every predetermined time period, and the multimedia resource that the terminal uploads to the server within the predetermined time period belongs to the multimedia resource to be released.

The identifier of the publisher of the multimedia resource to be released may be an account, a user name, and the like of the publisher in the media service platform.

S502. For each multimedia resource to be released, obtain historical behavior data of the publisher in the media service platform according to the identifier of the publisher of the multimedia resource to be released.

For the process of obtaining the historical behavior data of the publisher, refer to the related introduction of obtaining the historical behavior data of the user mentioned in the foregoing multimedia recommendation method, and details are not described herein.

S503: Determine, according to the historical behavior data of the publisher, a credit score model that reflects a mapping relationship between the historical behavior data and the credit score, and determine the credit score of the publisher.

Wherein, determining the credit score of the publisher based on the credit scoring model is an optional implementation manner, and in actual applications, the mapping relationship between the historical behavior data and the credit score may also be preset, and according to the history of the publisher. The behavior data and the mapping relationship determine the publisher's credit score.

The process of specifically determining the credit score of the publisher is similar to the process of determining the credit score of the historical visiting user or the target user mentioned in the previous embodiment, and details are not described herein.

The training process of the credit scoring model may also be: first obtaining a third-party credit score of a plurality of users as training samples and historical behavior data of a plurality of users; and then, according to multiple users' third-party credit scores and multiple The historical behavior data of the user and the preset regression model are subjected to regression analysis to obtain the scoring model. The specific process can also be related to the training credit scoring model in the previous embodiment, and will not be described here.

It can be understood that FIG. 5 is that the server obtains the historical behavior data of the publisher in real time after obtaining the identifier of the publisher, and determines the credit score of the publisher based on the historical behavior data as an example. Introduction. However, in actual applications, the credit scores of different users may also be calculated and stored in advance, and after obtaining the identifier of the publisher, the server may directly query the credit score of the publisher from the stored credit scores of different users.

S504. Sort the multimedia resources to be published according to the highest priority of the publishers of the multimedia resources to be released.

Among them, the higher the credit score of the publisher, the higher the ranking of the multimedia resources.

Due to the large number of multimedia resources that need to be released in the same period, the previous research and analysis show that the higher the publisher's credit score, the richer the multimedia resources of the publisher are, and the quality is higher. It can be seen that the higher the publisher's credit score, the higher the quality of the multimedia resources published by the publisher. Therefore, in order to enable high-quality multimedia resources to be preferentially recommended, a plurality of multimedia resources to be published may be sorted based on the publisher's credit score from high to low.

In some possible implementation manners of the present application, for each multimedia resource to be released, after determining the credit score of the publisher of the multimedia resource, the credit score of the publisher of the multimedia resource to be released and the quality of the multimedia resource may also be used. The mapping relationship between the scores determines the quality score of the multimedia resource to be published; wherein, in the mapping relationship, the higher the credit score of the publisher, the higher the quality score of the multimedia resource. Correspondingly, the multimedia resources to be distributed may be sorted according to the highest to lowest quality score of the multimedia resources to be released.

The mapping relationship between the credit score of the publisher of the multimedia resource and the quality score of the multimedia resource may have multiple possibilities. For example, the credit score of the publisher may be directly regarded as the quality score of the multimedia resource; for example, the publisher The credit score can be linearly related to the quality score of the multimedia resource.

Of course, the mapping relationship may have other possibilities, and is not limited herein.

S505. Determine a sorting order of the multimedia resources to be released, and determine a recommended order of the multimedia resources to be released to the terminal.

In this embodiment, after determining the sort order of the multimedia resources to be released, the sort order of the to-be-published multimedia resources may be determined as the recommended order of recommending the newly released multimedia resources to the terminal.

For example, the recommended order of the multimedia resources to be released may be pushed to the currently logged-in terminal. Or, when the page presentation request of the terminal is received, the multimedia presentation page is returned to the terminal according to the recommendation order of the multimedia resource to be published, so that the multimedia resource to be published is recommended according to the recommendation order in the multimedia presentation page.

It can be seen that, in this embodiment, after determining at least one to-be-published multimedia resource and the identifier of the publisher corresponding to each to-be-published multimedia resource, the publisher's credit score may be obtained, due to the credit score of the publisher of the multimedia resource. The quality of the multimedia resource can be reflected. Therefore, after sorting the plurality of multimedia resources in order of the publisher's credit score from high to low, the ranking of the high-quality multimedia resources is more advanced. The sorting order of the multimedia resources is used as a recommendation order for recommending multimedia resources to the terminal, which is beneficial to recommending high-quality multimedia resources issued by the publisher to the terminal, thereby improving the efficiency of the user positioning the high-quality multimedia resources. It is beneficial to improve the enthusiasm of users to access multimedia resources, thereby improving the click-through rate of multimedia resources and promoting the rapid dissemination of multimedia resources.

A multimedia recommendation device of the present application will be described below.

FIG. 6 is a schematic structural diagram of an embodiment of a multimedia recommendation apparatus according to the present application. The apparatus of this embodiment may include:

The request receiving unit 601 is configured to receive a page presentation request sent by the terminal, where the page presentation request is used to request the multimedia presentation page, and determine at least one multimedia resource to be recommended;

The quality determining unit 602 is configured to determine, for each multimedia resource to be recommended, the credit rating of the historical access user of the multimedia resource to be recommended, the number of historical access users of the multimedia resource to be recommended, and a preset scoring rule. The quality score of the multimedia resource, wherein the preset scoring rule is that the number of historical access users of the multimedia resource is positively correlated with the quality score of the multimedia resource, and/or the credit score of the historical access user of the multimedia resource and the quality score of the multimedia resource are positive Related

The resource sorting unit 603 is configured to sort the recommended multimedia resources according to the highest to lowest quality score of the multimedia resources to be recommended;

The resource presentation unit 604 is configured to feed back the multimedia presentation page to the terminal, and recommend the multimedia resource to be recommended according to the order of the multimedia resources to be recommended in the multimedia presentation page.

Quality determination unit, including:

a user determining unit, configured to determine a historical user set that accesses the multimedia resource to be recommended, where the historical access set includes at least one historical access user of the multimedia resource to be recommended;

a credit determining unit, configured to separately obtain a credit score of each historical access user of the multimedia resource to be recommended;

a quality scoring unit, configured to use, according to the historical access user of the to-be-recommended multimedia resource in the historical user set, and the historical access user's credit score of each of the to-be-recommended multimedia resources, and use a quality scoring model that conforms to a preset scoring rule, Determine the quality score of the multimedia resource to be recommended.

The quality determining unit is configured to: query the quality score of the to-be-recommended multimedia resource from the stored quality score set, where the quality score of the to-be-recommended multimedia resource is based on the number of historical access users accessing the to-be-recommended multimedia resource And a credit score of the historical access user of each of the to-be-recommended multimedia resources, and the score determined by the quality scoring model conforming to the preset scoring rule.

The page presentation request carries the identifier of the target user who logs in to the server through the terminal;

The device also includes:

a credit scoring unit, configured to obtain a credit score of the target user according to the identifier of the target user;

a difference determining unit, configured to determine, for each multimedia resource to be recommended, a difference between a credit score of the target user and a quality score of the multimedia resource to be recommended;

Resource sorting unit, including:

a first sorting unit, configured to sort the recommended multimedia resources according to an absolute value of the difference values from small to large;

a second sorting unit, configured to: at least two multimedia resources to be recommended having the same absolute value of the difference are used as a sorting set in the sorting process, according to the order of the quality scores of the multimedia resources to be recommended in the sorting set from high to low, Sort the multimedia resources to be recommended in the collection.

Credit score unit, including:

a historical data obtaining unit, configured to query historical behavior data of the target user according to the identifier of the target user;

The information score determining unit is configured to determine the credit score of the target user based on the historical behavior data of the target user and using a credit scoring model reflecting a mapping relationship between the historical behavior data and the credit score.

The credit scoring unit is specifically configured to: from the stored user credit score set according to the identity of the target user In the middle, check the credit score of the target user.

On the other hand, the embodiment of the present application further provides another multimedia recommendation device.

FIG. 7 is a schematic structural diagram of a multimedia recommendation device according to another application of the present application. The device may include:

The information acquiring unit 701 is configured to acquire at least one multimedia resource to be released, and an identifier of a publisher of each multimedia resource to be released;

The credit determining unit 702 is configured to determine, according to the identifier of the publisher, a credit score of the publisher for each multimedia resource to be released;

The resource sorting unit 703 is configured to sort the multimedia resources to be distributed according to the highest to lowest order of the credit scores of the publishers of the multimedia resources to be released;

The recommendation processing unit 704 is configured to determine a sorting order of the multimedia resources to be released as a recommendation order for recommending the multimedia resources to be released to the terminal.

Resource sorting unit, including:

a quality scoring unit, configured to determine, according to a mapping relationship between a credit score of a publisher of the multimedia resource to be released and a quality score of the multimedia resource, for each multimedia resource to be released, a quality score of the multimedia resource to be released;

The sorting sub-unit is configured to sort the multimedia resources to be distributed according to the highest to lowest quality score of the multimedia resources to be released.

Credit determination unit, including:

a data obtaining unit, configured to obtain historical behavior data of the publisher in the media service platform according to the identifier of the publisher;

The score analysis unit is configured to determine the credit score of the publisher according to the historical behavior data of the publisher and using a credit score model reflecting the mapping relationship between the historical behavior data and the credit score.

The credit determining unit is specifically configured to: the credit determining unit is specifically configured to:

According to the identifier of the publisher, the credit score of the publisher is queried from the stored set of user credit scores, wherein the publisher's credit score is based on the historical behavior data of the publisher, and the mapping between the historical behavior data and the credit score is reflected. The score of the relationship is determined by the credit scoring model.

The embodiment of the invention further provides a server, which may include the first multimedia mentioned above Recommended device.

8 is a block diagram showing the hardware structure of the server. Referring to FIG. 8, the server 800 may include: a processor 801, a communication interface 802, a memory 803, and a communication bus 804;

The processor 801, the communication interface 802, and the memory 803 complete communication with each other through the communication bus 804.

The communication interface 802 can be an interface of the communication module, such as an interface of the GSM module;

a processor 801, configured to execute a program;

a memory 803, configured to store a program;

The program can include program code, the program code including computer operating instructions.

The processor 801 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 803 may include a high speed RAM memory and may also include a non-volatile memory such as at least one disk memory.

Among them, the program can be specifically used to:

Receiving a page presentation request sent by the terminal, where the page presentation request is for requesting the multimedia presentation page;

Determining at least one multimedia resource to be recommended;

Determining the quality score of the multimedia resource to be recommended, according to the credit score of the historical access user of the multimedia resource to be recommended, the number of historical access users of the multimedia resource to be recommended, and the preset scoring rule, The preset scoring rule is that the number of historical access users of the multimedia resource is positively correlated with the quality score of the multimedia resource, and/or the credit score of the historical access user of the multimedia resource is positively correlated with the quality score of the multimedia resource;

Sorting the recommended multimedia resources according to the highest to lowest quality scores of the multimedia resources to be recommended;

The multimedia presentation page is fed back to the terminal, and the multimedia resource to be recommended is recommended according to the order of the multimedia resources to be recommended in the multimedia presentation page.

Correspondingly, determining a quality score of the to-be-recommended multimedia resource according to the credit score of the historical access user of the to-be-recommended multimedia resource, the number of historical access users of the to-be-recommended multimedia resource, and a preset scoring rule, including:

Determining a historical user set that accesses the to-be-recommended multimedia resource, where the historical access set includes at least one historical access user of the multimedia resource to be recommended;

Obtaining a credit score of each historical access user of the multimedia resource to be recommended separately;

Determining the to-be-recommended multimedia resource according to the number of historical access users of the to-be-recommended multimedia resource in the historical user set and the credit score of each historical access user of the to-be-recommended multimedia resource, and using a quality scoring model that conforms to a preset scoring rule Quality rating.

Correspondingly, determining a quality score of the to-be-recommended multimedia resource according to the credit score of the historical access user of the to-be-recommended multimedia resource, the number of historical access users of the to-be-recommended multimedia resource, and a preset scoring rule, including:

The quality score of the to-be-recommended multimedia resource is obtained from the stored quality score set, wherein the quality score of the to-be-recommended multimedia resource is based on the number of historical access users accessing the to-be-recommended multimedia resource and each of the multimedia to be recommended The history of the resource accesses the user's credit score and uses the quality scoring model that meets the pre-set scoring rules to determine the score.

Correspondingly, the page presentation request carries the identifier of the target user who logs in to the server through the terminal;

Before sorting the recommended multimedia resources according to the highest to lowest quality scores of the multimedia resources to be recommended, the following are:

Obtain a credit score of the target user according to the identifier of the target user;

Determining, by each of the multimedia resources to be recommended, a difference between the credit score of the target user and the quality score of the multimedia resource to be recommended;

Sort the recommended multimedia resources according to the highest to lowest quality scores of the multimedia resources to be recommended, including:

Sorting the recommended multimedia resources according to the order of the absolute values of the differences;

In the sorting process, at least two multimedia resources to be recommended having the same absolute value of the difference are used as a sorting set, and the multimedia resources to be recommended in the sorting set are sorted according to the quality score of the multimedia resources to be recommended in the sorting set from high to low. Sort.

Correspondingly, according to the identity of the target user, the credit score of the target user is obtained, including:

Query the historical behavior data of the target user according to the identifier of the target user;

Based on the historical behavior data of the target user, and using a credit scoring model reflecting the mapping relationship between the historical behavior data and the credit score, the credit score of the target user is determined.

Correspondingly, the credit scoring model reflecting the mapping relationship between historical behavior data and credit scores is trained as follows:

Obtaining a third party credit score of a plurality of users as a training sample and historical behavior data of a plurality of users;

According to the third-party credit rating of multiple users and the historical behavior data of multiple users, and the regression model of the preset regression analysis, a credit scoring model reflecting the mapping relationship between historical behavior data and credit score is obtained.

Correspondingly, according to the identity of the target user, the credit score of the target user is obtained, including:

According to the identifier of the target user, the credit score of the target user is queried from the stored set of user credit scores.

The embodiment of the present application further provides a storage medium for storing program code, and the program code is used to execute the first multimedia recommendation method provided in the foregoing embodiment.

The embodiment of the present application further provides a computer program product including instructions, which when executed on a server, causes the server to execute the first multimedia recommendation method provided in the foregoing embodiment. The embodiment of the invention further provides a server, which may include the second multimedia recommendation device.

9 is a block diagram showing the hardware structure of the server. Referring to FIG. 9, the server 900 may include: a processor 901, a communication interface 902, a memory 903, and a communication bus 904;

The processor 901, the communication interface 902, and the memory 903 complete communication with each other through the communication bus 904.

The communication interface 902 can be an interface of the communication module, such as an interface of the GSM module;

a processor 901, configured to execute a program;

a memory 903, configured to store a program;

The program can include program code, the program code including computer operating instructions.

The processor 901 may be a central processing unit CPU, or an Application Specific Integrated Circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present invention.

The memory 903 may include a high speed RAM memory and may also include a non-volatile memory such as at least one disk memory.

Among them, the program can be specifically used to:

Obtaining at least one multimedia resource to be released, and an identifier of a publisher of each multimedia resource to be released;

For each multimedia resource to be released, the publisher's credit score is determined according to the identifier of the publisher;

Sorting the multimedia resources to be distributed according to the highest priority of the publisher's credit scores of the multimedia resources to be released;

The sort order of the multimedia resources to be released is determined as a recommended order for recommending the multimedia resources to be released to the terminal.

Correspondingly, the multimedia resources to be published are sorted according to the highest priority of the publishers of the multimedia resources to be released, including:

For each multimedia resource to be released, determining a quality score of the multimedia resource to be released according to a mapping relationship between a credit score of the publisher of the multimedia resource to be released and a quality score of the multimedia resource;

The multimedia resources to be published are sorted according to the highest to lowest quality scores of the multimedia resources to be released.

Correspondingly, the publisher’s credit score is determined based on the publisher’s identity, including:

Obtain historical behavior data of the publisher in the media service platform according to the identifier of the publisher;

The publisher's credit score is determined based on the publisher's historical behavior data and a credit scoring model that reflects the mapping relationship between historical behavior data and credit scores.

Correspondingly, the publisher’s credit score is determined based on the publisher’s identity, including:

According to the identifier of the publisher, the credit score of the publisher is queried from the stored set of user credit scores, wherein the publisher's credit score is based on the historical behavior data of the publisher, and the mapping between the historical behavior data and the credit score is reflected. The score of the relationship is determined by the credit scoring model.

Correspondingly, the credit scoring model reflecting the mapping relationship between historical behavior data and credit scores is trained as follows:

Obtaining a third party credit score of a plurality of users as a training sample and historical behavior data of a plurality of users;

According to the third-party credit rating of multiple users and the historical behavior data of multiple users, and the regression model of the preset regression analysis, a credit scoring model reflecting the mapping relationship between historical behavior data and credit score is obtained.

The embodiment of the present application further provides a storage medium for storing program code, and the program code is used to execute the first multimedia recommendation method provided in the foregoing embodiment.

The embodiment of the present application further provides a computer program product including instructions, which when executed on a server, causes the server to execute the first multimedia recommendation method provided in the foregoing embodiment.

It should be noted that each embodiment in the specification is described in a progressive manner, and each embodiment focuses on differences from other embodiments, and the same similar parts between the embodiments are referred to each other. can. For the device type embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the description of the method embodiment.

Finally, it should also be noted that in this context, relational terms such as first and second are used merely to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply these entities. There is any such actual relationship or order between operations. Furthermore, the term "comprises" or "comprises" or "comprises" or any other variations thereof is intended to encompass a non-exclusive inclusion, such that a process, method, article, or device that comprises a plurality of elements includes not only those elements but also Other elements, or elements that are inherent to such a process, method, item, or device. An element defined by the phrase "comprising a ..." without further limitation does not exclude the presence of additional identical elements in the process, method, article, or device that comprises the element.

The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments are obvious to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the present invention is not to be limited to the embodiments shown herein, but the scope of the invention is to be accorded

The above is only a preferred embodiment of the present invention, and it should be noted that those skilled in the art can also make several improvements and retouchings without departing from the principles of the present invention. These improvements and retouchings should also be considered. It is the scope of protection of the present invention.

Claims (16)

A multimedia recommendation method, including: The server receives a page presentation request sent by the terminal, where the page presentation request is used to request a multimedia presentation page; Determining, by the server, at least one multimedia resource to be recommended; For each of the multimedia resources to be recommended, the server determines the multimedia resource to be recommended according to the credit score of the historical access user of the multimedia resource to be recommended, the number of historical access users of the multimedia resource to be recommended, and a preset scoring rule. Quality score, wherein the preset scoring rule is that the number of historical access users of the multimedia resource is positively correlated with the quality score of the multimedia resource, and/or the credit score of the historical access user of the multimedia resource is positive with the quality score of the multimedia resource. Related The server sorts the to-be-recommended multimedia resources according to the highest to lowest quality scores of the to-be-recommended multimedia resources; The server feeds back the multimedia presentation page to the terminal, and recommends the multimedia resource to be recommended according to the sort order of the multimedia resources to be recommended in the multimedia presentation page. The multimedia recommendation method according to claim 1, wherein the server determines the multimedia to be recommended according to the credit score of the historical access user of the multimedia resource to be recommended, the number of historical access users of the multimedia resource to be recommended, and a preset scoring rule. The quality score of the resource, including: Determining, by the server, a historical user set that accesses the to-be-recommended multimedia resource, where the historical access set includes at least one historical access user of the multimedia resource to be recommended; Obtaining, by the server, a credit score of each historical access user of the multimedia resource to be recommended; The server determines, according to the historical access user number of the to-be-recommended multimedia resource in the historical user set, and the historical credit user's credit score of each of the to-be-recommended multimedia resources, and determines the quality scoring model according to the preset scoring rule. The quality score of the multimedia resource to be recommended. The multimedia recommendation method according to claim 1, wherein the server determines the multimedia to be recommended according to the credit score of the historical access user of the multimedia resource to be recommended, the number of historical access users of the multimedia resource to be recommended, and a preset scoring rule. The quality score of the resource, including: The server queries the quality of the to-be-recommended multimedia resource from the stored quality score set a score, wherein the quality score of the to-be-recommended multimedia resource is based on the number of historical access users accessing the to-be-recommended multimedia resource and the credit score of each historical access user of the to-be-recommended multimedia resource, and using a predetermined scoring rule The score determined by the quality scoring model. The multimedia recommendation method according to claim 1, wherein the page presentation request carries an identifier of a target user who logs in to the server through the terminal; Before the server sorts the to-be-recommended multimedia resources in the order of the highest quality of the multimedia resources to be recommended, the server further includes: The server acquires a credit score of the target user according to the identifier of the target user; For each of the multimedia resources to be recommended, the server determines a difference between a credit score of the target user and a quality score of the multimedia resource to be recommended; The server sorts the to-be-recommended multimedia resources according to the highest to lowest quality scores of the to-be-recommended multimedia resources, including: The server sorts the multimedia resources to be recommended according to an order in which the absolute values of the differences are small to large; In the sorting process, the server uses at least two to-be-recommended multimedia resources whose absolute values are the same as the sorting set, according to the order of the quality scores of the multimedia resources to be recommended in the sorted set from high to low, The multimedia resources to be recommended in the sorting set are sorted. The multimedia recommendation method according to claim 4, the server acquiring the credit score of the target user according to the identifier of the target user, including: The server queries historical behavior data of the target user according to the identifier of the target user; The server determines the credit score of the target user based on the historical behavior data of the target user and using a credit scoring model reflecting a mapping relationship between the historical behavior data and the credit score. The multimedia recommendation method according to claim 5, wherein the credit score model reflecting the mapping relationship between the historical behavior data and the credit score is trained as follows: The server acquires a third party credit score of a plurality of users as a training sample and historical behavior data of the plurality of users; The server performs regression analysis according to the third-party credit score of the plurality of users and historical behavior data of the plurality of users, and a preset regression model, to obtain the historical behavior data and the reflected A credit scoring model for mapping relationships between credit scores. The multimedia recommendation method according to claim 4, the server acquiring the credit score of the target user according to the identifier of the target user, including: The server queries the credit score of the target user from the stored set of user credit scores according to the identifier of the target user. A multimedia recommendation method, including: Obtaining, by the server, at least one multimedia resource to be released, and an identifier of a publisher of each of the to-be-published multimedia resources; Determining, by the server, a credit score of the publisher according to the identifier of the publisher for each of the to-be-published multimedia resources; The server sorts the to-be-published multimedia resources in descending order of the credit scores of the publishers of the to-be-published multimedia resources; And determining, by the server, a sorting order of the to-be-published multimedia resources, to recommend a recommended order of the to-be-published multimedia resources to the terminal. The multimedia recommendation method according to claim 8, wherein the server sorts the to-be-published multimedia resources according to the highest priority of the publisher's credit score of the to-be-published multimedia resource, including: For each of the to-be-published multimedia resources, the server determines a quality score of the to-be-published multimedia resource according to a mapping relationship between a credit score of a publisher of the to-be-published multimedia resource and a quality score of the multimedia resource; The server sorts the to-be-published multimedia resources according to the highest to lowest quality score of the to-be-published multimedia resources. The multimedia recommendation method according to claim 8, wherein the server determines the credit score of the publisher according to the identifier of the publisher, including: The server acquires historical behavior data of the publisher in the media service platform according to the identifier of the publisher; The server determines the credit score of the publisher according to the historical behavior data of the publisher and using a credit scoring model reflecting a mapping relationship between the historical behavior data and the credit score. The multimedia recommendation method according to claim 8, wherein said server is based on said publisher The identification of the publisher’s credit score, including: The server queries the publisher's credit score from the stored user credit score set according to the identifier of the publisher, where the publisher's credit score is based on the publisher's historical behavior data, and A score determined using a credit scoring model reflecting a mapping relationship between the historical behavior data and a credit score. The multimedia recommendation method according to claim 10 or 11, wherein the credit score model reflecting the mapping relationship between the historical behavior data and the credit score is trained as follows: The server acquires a third party credit score of a plurality of users as a training sample and historical behavior data of the plurality of users; The server performs regression analysis according to the third-party credit score of the plurality of users and historical behavior data of the plurality of users, and a preset regression model, to obtain a mapping relationship between the reflected historical behavior data and the credit score. Credit score model. A server, the server comprising: a processor and a memory; The memory is configured to store program code and transmit the program code to the processor; The processor is configured to invoke an instruction in a memory to perform the multimedia recommendation method according to any one of claims 1-7. A server, the server comprising: a processor and a memory; The memory is configured to store program code and transmit the program code to the processor; The processor is configured to invoke an instruction in a memory to perform the multimedia recommendation method according to any one of claims 8-12. A storage medium for storing program code for performing the multimedia recommendation method of any one of claims 1-7. A storage medium for storing program code for performing the multimedia recommendation method of any one of claims 8-12.

Images