阅读说明：本技术 广告召回方法和系统 (Advertisement recalling method and system ) 是由 吴安新 牛心怡 林婧 于 2019-10-25 设计创作，主要内容包括：本申请实施例提供了一种广告召回方法,所述广告召回方法包括：获取目标视频的目标视频标签；计算所述目标视频标签与映射表中各个视频标签之间的相关度；从所述多个视频标签中筛选出相关度大于预设阈值的多个相关视频标签；将所述多个相关视频标签映射的至少一个兴趣标签与所述目标视频进行关联操作,以根据所述至少一个兴趣标签为所述目标视频执行广告召回操作。本申请实施例提供的技术方案,能够提高广告召回效率和广告召回速度。(The embodiment of the application provides an advertisement recalling method, which comprises the following steps: acquiring a target video label of a target video; calculating the correlation degree between the target video label and each video label in a mapping table; screening out a plurality of related video tags of which the correlation degrees are greater than a preset threshold value from the plurality of video tags; and associating at least one interest tag mapped by the plurality of related video tags with the target video so as to execute advertisement recalling operation for the target video according to the at least one interest tag. The technical scheme provided by the embodiment of the application can improve the advertisement recall efficiency and the advertisement recall speed.)

1. An advertisement recall method, the method comprising:

acquiring a target video label of a target video;

calculating the correlation degree between the target video label and each video label in a mapping table, wherein the mapping table comprises a plurality of interest labels and a plurality of video labels, and mapping relations exist between the interest labels and the video labels;

screening out a plurality of related video tags of which the correlation degrees are greater than a preset threshold value from the plurality of video tags;

and associating at least one interest tag mapped by the plurality of related video tags with the target video so as to execute advertisement recalling operation for the target video according to the at least one interest tag.

2. The advertisement recall method of claim 1, further comprising:

training based on a corpus to obtain a word vector model;

the corpus is composed of a plurality of video participles, and the video participles comprise a plurality of video labels and a plurality of title participles of a plurality of videos.

3. The advertisement recall method of claim 2 wherein the relevancy comprises a cosine similarity; calculating the correlation degree between the target video label and each video label in the mapping table, wherein the calculation comprises the following steps:

converting the target video label into a target video label word vector through the word vector model;

and calculating the cosine similarity between the target video label word vector and the video label word vectors of the video labels respectively.

4. The advertisement recall method of claim 1, further comprising:

obtaining the plurality of interest tags;

performing word segmentation operation on each interest tag to obtain a plurality of groups of core words, wherein each group of core words corresponds to one interest tag;

calculating the correlation degree between each core word in each group of core words and each video label in the corpus;

acquiring a plurality of video label sets according to the correlation degree between each core word in each group of core words and each video label in the corpus; and

and establishing a mapping relation between each video label set and the corresponding interest label to construct the mapping table.

5. The advertisement recall method of claim 1 wherein associating at least one interest tag of the plurality of related video tag maps with the target video comprises:

obtaining a plurality of target interest tags through the plurality of related video tags;

determining the correlation degree value between each related video tag and the target video tag as the correlation degree value between the corresponding target interest tag and the target video; and

and mounting the plurality of target interest tags and the corresponding relevance numerical value of each target interest tag for the target video.

6. The advertisement recall method of claim 5, further comprising:

receiving a page access request from a client terminal, wherein the page access request is used for requesting to acquire a playing page of the target video;

acquiring partial advertisements from an advertisement library according to the plurality of target interest tags mounted on the target video and the corresponding relevance numerical value of each target interest tag; and

returning the partial advertisement to the client terminal.

7. The advertisement recalling method according to claim 6, wherein the step of obtaining partial advertisements from an advertisement library according to the plurality of target interest tags mounted on the target video and the corresponding relevance value of each target interest tag comprises:

respectively comparing a correlation threshold value with a correlation numerical value corresponding to each target interest tag, wherein the correlation threshold value is preset and adjustable;

selecting a part of target interest tags from the plurality of target interest tags according to the comparison result; and

and acquiring the partial advertisement from the advertisement library according to the partial target interest tag.

8. An advertisement recall system, comprising:

the acquisition module is used for acquiring a target video label of a target video;

the calculation module is used for calculating the correlation degree between the target video label and each video label in a mapping table, the mapping table comprises a plurality of interest labels and a plurality of video labels for video identification, and mapping relations exist between the interest labels and the video labels;

the screening module is used for screening out a plurality of related video tags of which the correlation degrees are greater than a preset threshold value from the plurality of video tags;

and the association module is used for performing association operation on at least one interest tag mapped by the plurality of related video tags and the target video so as to execute advertisement recall operation for the target video according to the at least one interest tag.

9. A computer device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor when executing the computer program is configured to implement the steps of the advertisement recall method of any one of claims 1-7.

10. A computer readable storage medium having stored therein a computer program executable by at least one processor to cause the at least one processor to perform the steps of the advertisement recall method of any one of claims 1-7.

Technical Field

The embodiment of the application relates to the technical field of internet, in particular to an advertisement recalling method, system, equipment and computer readable storage medium.

Background

With the development of internet services, people are beginning to perform more and more life entertainment, various transactions and the like by means of a network platform. How to expose goods or service information to customers or potential customers through a network becomes a concern for all parties.

Among the widely known solutions is by placing advertisements on web pages. Specifically, an advertiser puts an advertisement to an advertisement platform in a bidding mode, and the advertisement platform adds the advertisement put in the advertisement pool. The advertisement platform will screen relevant advertisements from the advertisement pool according to the current video information watched by the user, the interest of the user, the targeting of the advertiser, and other conditions, and expose the screened relevant advertisements to the user, which may be called advertisement recall. It is understood that when the number of advertisements in the advertisement pool is very large, which means that a large number of advertisements need to be recalled, the advertisement recall speed and efficiency are very low.

Disclosure of Invention

An object of the embodiments of the present application is to provide an advertisement recall method, system, computer device and computer-readable storage medium, which are used to solve the technical problem that the advertisement recall speed and efficiency are very low.

One aspect of an embodiment of the present application provides an advertisement recall method, including: acquiring a target video label of a target video; calculating the correlation degree between the target video label and each video label in a mapping table, wherein the mapping table comprises a plurality of interest labels and a plurality of video labels for video identification, and mapping relations exist between the plurality of interest labels and the plurality of video labels; screening out a plurality of related video tags of which the correlation degrees are greater than a preset threshold value from the plurality of video tags; and associating at least one interest tag mapped by the plurality of related video tags with the target video so as to execute advertisement recalling operation for the target video according to the at least one interest tag.

Optionally, the method further includes: training based on a corpus to obtain a word vector model; the corpus is composed of a plurality of video participles, and the video participles comprise a plurality of video labels and a plurality of title participles of a plurality of videos.

Optionally, the correlation includes cosine similarity; calculating the correlation degree between the target video label and each video label in the mapping table, wherein the calculation comprises the following steps: converting the target video label into a target video label word vector through the word vector model; and calculating the cosine similarity between the target video label word vector and the video label word vectors of the video labels respectively.

Optionally, the method further includes: obtaining the plurality of interest tags; performing word segmentation operation on each interest tag to obtain a plurality of groups of core words, wherein each group of core words corresponds to one interest tag; calculating the correlation degree between each core word in each group of core words and each video label in the corpus; acquiring a plurality of video label sets according to the correlation degree between each core word in each group of core words and each video label in the corpus; and establishing a mapping relation between each video label set and the corresponding interest label to construct the mapping table.

Optionally, associating at least one interest tag mapped to the plurality of related video tags with the target video, including: obtaining a plurality of target interest tags through the plurality of related video tags; determining the correlation degree value between each related video tag and the target video tag as the correlation degree value between the corresponding target interest tag and the target video; and mounting the plurality of target interest tags and the corresponding relevance numerical value of each target interest tag for the target video.

Optionally, the method further includes: receiving a page access request from a client terminal, wherein the page access request is used for requesting to acquire a playing page of the target video; acquiring partial advertisements from an advertisement library according to the plurality of target interest tags mounted on the target video and the corresponding relevance numerical value of each target interest tag; and returning the partial advertisement to the client terminal.

Optionally, obtaining a part of advertisements from an advertisement library according to the multiple target interest tags mounted on the target video and the relevance value corresponding to each target interest tag, where the obtaining includes: respectively comparing a correlation threshold value with a correlation numerical value corresponding to each target interest tag, wherein the correlation threshold value is preset and adjustable; selecting a part of target interest tags from the plurality of target interest tags according to the comparison result; and acquiring the partial advertisement from the advertisement library according to the partial target interest tag.

An aspect of an embodiment of the present application further provides an advertisement recall method, including: the acquisition module is used for acquiring a target video label of a target video; the calculation module is used for calculating the correlation degree between the target video label and each video label in a mapping table, the mapping table comprises a plurality of interest labels and a plurality of video labels for video identification, and mapping relations exist between the interest labels and the video labels; the screening module is used for screening out a plurality of related video tags of which the correlation degrees are greater than a preset threshold value from the plurality of video tags; and the association module is used for performing association operation on at least one interest tag mapped by the plurality of related video tags and the target video so as to execute advertisement recall operation for the target video according to the at least one interest tag.

An aspect of the embodiments of the present application further provides a computer device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor is configured to implement the steps of the advertisement recall method described above when executing the computer program.

An aspect of the embodiments of the present application further provides a computer-readable storage medium, in which a computer program is stored, where the computer program is executable by at least one processor to cause the at least one processor to execute the steps of the advertisement recall method.

The advertisement recall method, the system, the equipment and the computer readable storage medium provided by the embodiment of the application can control the number of interest tags in the target video through the size of the preset threshold value, and further control the advertisement recall amount corresponding to the target video, so that the problem of mass advertisement recall is solved, and the advertisement recall efficiency and the advertisement recall speed are improved. In addition, the interest tags of the target videos are found through the target video tags, and then the advertisement recall operation is carried out through the found interest tags, so that the strong correlation between the video tags and the interest tags (such as advertisement categories) is effectively utilized, the retrieval range of the advertisement recall is effectively reduced, and the advertisement recall efficiency is improved.

Drawings

FIG. 1 schematically illustrates an environmental application diagram according to an embodiment of the present application;

FIG. 2 is a flow chart of an advertisement recall method according to an embodiment one of the present application;

FIG. 3 is a flow chart schematically illustrating an advertisement recall method according to a second embodiment of the present application;

FIG. 4 schematically illustrates a flow chart for building a mapping table;

FIG. 5 schematically illustrates a flow chart for calculating a correlation;

FIG. 6 is a flow chart schematically illustrating an advertisement recall method according to a third embodiment of the present application;

fig. 7 schematically shows a detailed flowchart of step S614;

FIG. 8 is a block diagram schematically illustrating an advertisement recall method according to a fourth embodiment of the present application; and

fig. 9 schematically shows a hardware architecture diagram of a computer device suitable for implementing the advertisement recall method according to an embodiment of the present application.

Detailed Description

In order to make the objects, technical solutions and advantages of the embodiments of the present application more clearly understood, the embodiments of the present application are described in further detail below with reference to the accompanying drawings and the embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the embodiments of the application and are not intended to limit the embodiments of the application. All other embodiments obtained by a person of ordinary skill in the art based on the embodiments in the present application without any creative effort belong to the protection scope of the embodiments in the present application.

It should be noted that the description relating to "first", "second", etc. in the present invention is for descriptive purposes only and is not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one such feature. In addition, technical solutions between various embodiments may be combined with each other, but must be realized by a person skilled in the art, and when the technical solutions are contradictory or cannot be realized, such a combination should not be considered to exist, and is not within the protection scope of the present invention.

Fig. 1 schematically shows an environment application diagram according to an embodiment of the application. In an exemplary embodiment, the provider network 2 may be connected to a plurality of client terminals 10 through one or more networks 9.

In an exemplary embodiment, the content service 3 may be implemented as part of the provider network 2. In other embodiments, the content service 3 may be managed by a separate service provider than the service provider of the provider network 2. It should also be understood that the provider network 2 may provide additional content services separate from the content service 3.

The content service 3 may comprise a content streaming service, such as an internet protocol video streaming service. Content streaming services may be configured to distribute content via various transmission techniques. The content service 3 may be configured to provide content such as video, audio, text data, combinations thereof, and/or the like. The content may include content streams (e.g., video streams, audio streams, information streams), content files (e.g., video files, audio files, text files), and/or other data.

The provider network 2 may be located in a data center, such as a single house, or distributed over different geographical locations (e.g., over several houses). The provider network 2 may provide services through one or more networks 9. Network 9 includes various network devices such as routers, switches, multiplexers, hubs, modems, bridges, repeaters, firewalls, proxy devices, and/or the like. The network 9 may include physical links, such as coaxial cable links, twisted pair cable links, fiber optic links, combinations thereof, and/or the like. The network 9 may include wireless links such as cellular links, satellite links, Wi-Fi links, and/or the like.

The provider network 2 may include an advertisement service 5. The ad service 5 may be configured with one or more ad pool(s) that may be used to deposit the advertiser's ads. Advertisements may consist of pictures, videos, text, etc.

The provider network 2 may also include a processing service 4. The processing service 4 may be configured to provide processing of various services, such as services of the provider network 2. Processing service 4 may include multiple processing nodes 8 (e.g., as servers). Multiple processing nodes 8 may process tasks associated with the ad service 5. The plurality of processing nodes 8 may be implemented as one or more computing devices, one or more processors, one or more virtual computing instances, combinations thereof, and/or the like.

The plurality of processing nodes 8 may be implemented by one or more computing devices. One or more computing devices may include a virtualized compute instance. The virtualized computing instance may include a virtual machine, such as an emulation of a computer system, an operating system, a server, and so forth. The computing device may load the virtual machine based on a virtual image and/or other data that defines the particular software (e.g., operating system, dedicated application, server) used for emulation. As the demand for different types of processing services changes, different virtual machines may be loaded and/or terminated on one or more computing devices. A hypervisor may be implemented to manage the use of different virtual machines on the same computing device.

The plurality of processing nodes 8 may include nodes associated with providing a particular service (e.g., processing task). The nodes may be dedicated to providing these specific services. For example, the plurality of processing nodes 8 may implement receivers, content generators, combiners, transcoders, ad dispensers, and combinations thereof.

In an exemplary embodiment, multiple processing nodes 8 may process events submitted by multiple client terminals. The event may be associated with an advertisement placement, a content streaming upload, a content streaming download, or an online live broadcast, etc. In an exemplary embodiment, these services may be implemented as dedicated computing devices, dedicated processors, dedicated virtual machine instances, and the like. In other implementations, various different nodes may implement any functionality associated with these services.

A plurality of client terminals 10 may be configured to access content and services of the provider network 2. The plurality of client terminals 10 may include any type of computing device, such as a mobile device, a tablet device, a laptop computer, a computing station, a smart device (e.g., smart apparel, smart watch, smart speaker, smart glasses), a virtual reality headset, a gaming device, a set-top box, a digital streaming device, a robot, a vehicle mounted terminal, a smart television, a television box, an e-book reader, an MP3 (motion picture experts group audio layer III) player, an MP4 (motion picture experts group audio layer IV) player, and so forth.

Multiple client terminals 10 may be associated with one or more users. A single user may access the provider network 2 using one or more of a plurality of client terminals 10. Multiple client terminals 10 may travel to various locations and use different networks to access the provider network 2.

Illustratively, the client terminal 10 may include an application 12. The application 12 outputs (e.g., displays or presents) the content to the user. The content may include video, audio, commentary, textual data, advertisements, and/or the like.

The client terminal 10 may access an interface 16, the interface 16 allowing a user to provide relevant content to the provider network 2, such as allowing an advertiser to submit an advertisement or advertisement keyword, etc., allowing an UP host to submit a video or video description information (such as a video tag, etc.), etc., or allowing a user to submit a search keyword, access request, etc. The UP master, which may refer to a video uploader. In some embodiments, the interface 16 may be implemented as part of the application 12.

The content service 3, processing service 4, advertising service 5, etc. described above may be implemented by one or more computer devices 20. It will be appreciated that computer device 20 may be used to execute any aspect of the computer described herein, for example, to implement the advertisement recall method described herein.

Example one

Fig. 2 schematically shows a flowchart of an advertisement recall method according to a first embodiment of the present application. It is to be understood that the flow charts in the embodiments of the present application are not used to limit the order of executing the steps.

As shown in fig. 2, the advertisement recall method may include steps S200 to S206, in which:

and step S200, acquiring a target video label of the target video.

The target video tag is used for identifying the video content of the target video, for example, if the target video is a makeup video, the target video tag may be "makeup", "makeup course", "summer makeup", or the like.

The interface 16 of the client terminal 10 may be configured to include the following: a video upload interface, a video description interface, etc. The video description interface may be a drop-down menu-type interface, a text-entry type interface, or the like. The pull-down menu type interface is preset with various options for the UP main selection. And the text input type interface is used for UP main input.

When the UP host uploads the target video to the computer device 20 through the client terminal 10, the video description information associated with the target video, such as a video title and at least one target video tag, can be simultaneously input or selected. For example, if the target video is a game video, a video tag such as "online game", "dead survival", etc. may be selected or input; if the target video is a makeup video, video tags such as "makeup", "makeup course", "summer makeup" and the like may be selected or input. Of course, the video description information associated with the target video may also be modified by the operator.

After the target video and the video description information associated with the target video are provided to the computer device 20, the computer device 20 extracts the target video tag from the video description information to obtain at least one interest tag according to the target video tag.

Step S202, calculating the correlation degree between the target video label and each video label in the mapping table.

The mapping table comprises a plurality of interest tags and a plurality of video tags, and mapping relations exist between the interest tags and the video tags. As shown in the following table:

interest tag	Video label
		A	a1、a2、a3、…
B	b1、b2、b3、…
		C	c1、c2、c3、…
D	d1、d2、d3、…
		…	…

It will be appreciated that there should be some correlation between interest tags and corresponding video tags, such as having similar semantics or having the same function. Also, the same video tag may be mapped to one or more interest tags.

And the interest tags are obtained by configuration according to various commercial interest categories corresponding to various advertisements. The plurality of interest tags can be classified into a plurality of levels of interest tags according to business interest categories, sub-categories and the like, for example, the level of interest tags are education, and the level of interest tags are Japanese education, English education and the like. In some embodiments, each interest tag may have an advertisement keyword.

A plurality of video tags, which may be derived from a corpus, the corpus consisting of at least: video tags (selected or entered by a user or a worker) for respective videos, and a plurality of title segments obtained by segmenting the video titles of the respective videos.

It should be noted that the mapping table is configured in advance, and can be updated periodically to add a new video tag or delete an old video tag.

Step S204, a plurality of relevant video labels with the correlation degree larger than a preset threshold value are screened out from the plurality of video labels.

The predetermined threshold may be preset or may be dynamically set.

In an exemplary embodiment, the preset threshold is set to 0.95, and if the degree of correlation between the video tag/tags of the plurality of video tags and the target video tag is greater than 0.95, the video tag/tags are regarded as related video tags.

In an exemplary embodiment, the threshold is dynamically set to X, and if some of the plurality of video tags are associated with the target video tag by more than X, then these video tags are considered to be associated video tags. Wherein the related video tags are associated with a preset number (e.g., 10) of interest tags. It is understood that the setting of X is based on the ability to screen out 10 interest tags for the target video.

Step S206, at least one interest tag mapped by the related video tags is associated with the target video, so as to execute an advertisement recall operation for the target video according to the at least one interest tag.

That is, if one or more video tags among the plurality of video tags mapped by a certain interest tag are related video tags, the interest tag is associated with the target video, for example, the interest tag is mounted under the target video. It is understood that when the target video is tagged with the interest tag, the target video may recall the advertisement or advertisements associated with the interest tag.

The technical scheme of the first embodiment has at least the following technical effects:

one is as follows: the advertisement recall quantity corresponding to the target video can be controlled by controlling the quantity of the interest tags in the target video, so that the problem of mass advertisement recall is solved, and the advertisement recall efficiency and the advertisement recall speed are improved. The method comprises the following specific steps:

by setting a predetermined threshold, the amount of advertisement recalls may be controlled. The reasoning is as follows: by controlling the size of the preset threshold value, the number of related video tags can be controlled, and further the number of interest tags mounted under the target video is controlled. If the number of interest tags mounted on the target video is increased, the advertisement recall amount is increased; and if the interest tag data mounted by the target video is reduced, the advertisement recall amount is reduced.

The second step is as follows: and obtaining the interest label with high relevance with the target video by calculating the relevance between the target video label of the target video and the video label mounted under each interest label in the mapping table. The interest tags of the target video can be directly and quickly determined only by carrying out correlation calculation with the video tags mounted under the interest tags in the mapping table, and calculation with all the video tags at each time is not needed, so that the calculation resources and the calculation speed are effectively saved.

And thirdly: the interest tags of the target videos are found through the target video tags, and then advertisement recall operation is carried out through the found interest tags, so that the strong correlation between the video tags and the interest tags (such as advertisement categories) is effectively utilized, the retrieval range of advertisement recall is effectively reduced, the advertisement recall efficiency is improved, and the high recall accuracy is achieved.

Fourthly, the method comprises the following steps: in the advertisement recall method in the traditional scheme, when mass advertisements need to be recalled, the operation resources of click-through-rate (CTR) estimation can be greatly increased;

in contrast, according to the technical scheme of the first embodiment of the application, since the plurality of interest tags and the video tags having strong correlation with the interest tags are pre-configured in advance in the form of the mapping table, the retrieval range of advertisement recall is effectively narrowed, that is, the number of recalled advertisements is reduced, the click rate is improved, and the calculation resources of click rate estimation are reduced.

Example two

Fig. 3 schematically shows a flowchart of an advertisement recall method according to the second embodiment of the present application. As shown in fig. 3, the advertisement recall method may include steps S300 to S308, in which:

step S300, a corpus is constructed and a word vector model is obtained based on the corpus training.

The corpus is composed of a plurality of video segments including a plurality of video tags and a plurality of title segments of a plurality of videos.

Taking a certain content provider as an example, the computer device 20 may perform the following operations: (1) determining all videos or part of videos stored by the content provider as sample videos; (2) acquiring a video label of each sample video, and performing word segmentation operation on a video title of each sample video to obtain a title word segmentation corresponding to each sample video; (3) using each video label and each title word obtained in the step (2) to construct the corpus; (4) and training a word2vector algorithm according to each word segmentation in the corpus to obtain the word vector model.

Based on the word2vector algorithm, semantic information of each participle in the corpus can be fully utilized. And the video labels and the video label participles are used as the corpus materials of the word vector model obtained by training, so that the interference of a large amount of invalid information can be avoided, the training burden can be reduced, and a smaller participle vector space can be maintained.

Step S302, a mapping table is constructed.

The mapping table comprises a plurality of interest tags and a plurality of video tags, and mapping relations are formed between the interest tags and the video tags. As shown in fig. 4, the step of constructing the mapping table is as follows: step S400, obtaining the plurality of interest tags. Step S402, performing word segmentation operation on each interest tag to obtain a plurality of groups of core words, wherein each group of core words corresponds to one of the interest tags. Step S404, calculating the correlation between each core word in each group of core words and each video label in the corpus. Step S406, a plurality of video label sets are obtained according to the correlation between each core word in each group of core words and each video label in the corpus. Step S408, establishing a mapping relation between each video label set and the corresponding interest label to construct the mapping table.

Taking the example of a "men's shoe," the computer device 20 may perform the following operations: obtaining an interest label 'men's clothing and men's shoes'; performing word segmentation operation on the 'men' shoes through a word segmentation algorithm to obtain a plurality of core words, such as 'men' shoes and 'men' shoes; calculating the correlation degree of the 'men' clothing and each video label in the corpus, and calculating the correlation degree of the 'men' shoes and each video label in the corpus; screening 20 video labels with the maximum correlation degree with the male clothes/men's shoes from the corpus, wherein the 20 segmented words are called as the correlation video labels of the male clothes/men's shoes; and establishing a mapping relation between the man's shoes and the 20 related video tags, and recording the mapping relation into a mapping table. In some cases, some video labels which are high in playing amount and are very related to the man shoes can be recorded into the mapping table through manual operation.

Steps S300 and S302 are preparatory steps, which consist in implementing vectorization of video tags and providing a mapping table representing the mapping relationship between interest tags and video tags. On the basis of the above preparation steps, the interest tag associated with each video can be configured, and the configuration process is described below by taking the target video as an example.

Step S304, a target video label of the target video is obtained.

Step S306, calculating the correlation degree between the target video label and each video label in the mapping table.

In an exemplary embodiment, the correlation may be characterized by a cosine similarity, as shown in fig. 5, the step S306 may include steps S500 to S502: step S500, converting the target video label into a target video label word vector through the word vector model. Step S502, calculating Cosine Similarity (Cosine Similarity) between the target video tag word vector and the video tag word vectors of the video tags, respectively. It should be noted that other coefficients may also be applied to represent the Correlation between the target video tag and each video tag in the mapping table in the embodiment of the present application, for example, euclidean Distance (euclidean Distance), Manhattan Distance (Manhattan Distance), Minkowski Distance (Minkowski Distance), Pearson Correlation Coefficient (Pearson Correlation Coefficient), and the like.

Step S308, a plurality of relevant video labels with the correlation degree larger than a preset threshold value are screened out from the plurality of video labels.

Step S310, performing association operation on at least one interest tag mapped by the plurality of related video tags and the target video, so as to execute advertisement recall operation for the target video according to the at least one interest tag.

That is, if one or more video tags among the plurality of video tags mapped by a certain interest tag are related video tags, the interest tag is associated with the target video, for example, the interest tag is mounted under the target video. It is understood that when the target video is tagged with the interest tag, the target video may recall the advertisement or advertisements associated with the interest tag.

EXAMPLE III

Fig. 6 schematically shows a flowchart of an advertisement recall method according to a third embodiment of the present application. As shown in fig. 6, the advertisement recall method may include steps S600 to S608, in which:

step S600, a target video label of the target video is obtained.

Step S602, calculating a correlation between the target video tag and each video tag in a mapping table, where the mapping table includes a plurality of interest tags and a plurality of video tags for video identification, and there is a mapping relationship between the plurality of interest tags and the plurality of video tags.

Step S604, a plurality of relevant video tags with a correlation degree greater than a preset threshold are screened out from the plurality of video tags.

Step S606, a plurality of target interest tags are obtained through the plurality of related video tags.

Step S608, determining the correlation value between each related video tag and the target video tag as the correlation value between the corresponding target interest tag and the target video.

Step S610, mounting the multiple target interest tags and the corresponding relevance value of each target interest tag for the target video.

For example, the 10 interest tags with the highest relevance (called target interest tags) are mounted under the target video, and the relevance values of the 10 interest tags with the target video are (target interest tag #1, 0.95), (target interest tag #2, 0.91), …, (target interest tag #10, 0.7), respectively.

Step S612, receiving a page access request from the client terminal, where the page access request is used to request to obtain a playing page of the target video.

Step S614, according to the multiple target interest tags mounted on the target video and the corresponding relevance value of each target interest tag, obtaining partial advertisements from an advertisement library.

The higher the correlation value corresponding to the target interest tag is, the higher the probability that the corresponding advertisement is recalled is. It should be noted that, in some embodiments, the probability of the advertisement being recalled is not completely determined by the relevance value, and may be influenced by the creative size of the advertisement, the IP location of the client, the current time period, and other factors.

In an exemplary embodiment, the advertisement recalling may be performed according to all the target interest tags mounted by the target video, or may be performed according to the advertisement quantity of the advertisement pool or the operation burden of the computer device 20 by selecting a part of the target interest tags mounted by the target video. As shown in fig. 7, the step S614 may include steps S700 to S704: step S700, comparing a correlation threshold with the correlation value corresponding to each target interest tag, wherein the correlation threshold is preset and adjustable. Step S702, according to the comparison result, selecting a part of target interest tags from the plurality of target interest tags. Step S704, according to the partial target interest tag, obtaining the partial advertisement from the advertisement library.

Step S616, returning the partial advertisement to the client terminal.

The partial advertisement can be displayed in an advertisement display area of the playing page and displayed to the user, and can also be displayed to the user in a pop-up mode in a pop-up layer.

Example four

Fig. 8 is a block diagram schematically illustrating an advertisement recall method according to a fourth embodiment of the present application, which may be divided into one or more program modules, the one or more program modules being stored in a storage medium and executed by one or more processors to implement the embodiments of the present application. The program modules referred to in the embodiments of the present application refer to a series of computer program instruction segments capable of performing specific functions, and the following description will specifically describe the functions of each program module in the embodiments.

As shown in FIG. 8, the advertisement recall method 800 may include an acquisition module 810, a calculation module 820, a screening module 830, an association module 840, a recall module 850, a training module 860, and a mapping module 870, wherein:

an obtaining module 810, configured to obtain a target video tag of a target video.

A calculating module 820, configured to calculate a correlation between the target video tag and each video tag in a mapping table, where the mapping table includes a plurality of interest tags and a plurality of video tags for video identification, and a mapping relationship exists between the plurality of interest tags and the plurality of video tags.

The screening module 830 is configured to screen a plurality of relevant video tags, of which the relevance is greater than a preset threshold, from the plurality of video tags.

An associating module 840, configured to perform an associating operation on at least one interest tag mapped to the multiple related video tags and the target video, so as to perform an advertisement recalling operation for the target video according to the at least one interest tag.

Optionally, a training module 860 for: training based on a corpus to obtain a word vector model; the corpus is composed of a plurality of video participles, and the video participles comprise a plurality of video labels and a plurality of title participles of a plurality of videos.

Optionally, the correlation includes cosine similarity; a calculation module 820 further configured to: calculating the correlation degree between the target video label and each video label in the mapping table, wherein the calculation comprises the following steps: converting the target video label into a target video label word vector through the word vector model; and calculating the cosine similarity between the target video label word vector and the video label word vectors of the video labels respectively.

Optionally, the mapping module 870 is configured to: obtaining the plurality of interest tags; performing word segmentation operation on each interest tag to obtain a plurality of groups of core words, wherein each group of core words corresponds to one interest tag; calculating the correlation degree between each core word in each group of core words and each video label in the corpus; acquiring a plurality of video label sets according to the correlation degree between each core word in each group of core words and each video label in the corpus; and establishing a mapping relation between each video label set and the corresponding interest label to construct the mapping table.

Optionally, the associating module 840 is configured to: obtaining a plurality of target interest tags through the plurality of related video tags; determining the correlation degree value between each related video tag and the target video tag as the correlation degree value between the corresponding target interest tag and the target video; and mounting the plurality of target interest tags and the corresponding relevance numerical value of each target interest tag for the target video.

Optionally, the recall module 850 is configured to: receiving a page access request from a client terminal, wherein the page access request is used for requesting to acquire a playing page of the target video; acquiring partial advertisements from an advertisement library according to the plurality of target interest tags mounted on the target video and the corresponding relevance numerical value of each target interest tag; and returning the partial advertisement to the client terminal.

Optionally, the recall module 850 is further configured to: respectively comparing a correlation threshold value with a correlation numerical value corresponding to each target interest tag, wherein the correlation threshold value is preset and adjustable; selecting a part of target interest tags from the plurality of target interest tags according to the comparison result; and acquiring the partial advertisement from the advertisement library according to the partial target interest tag.

EXAMPLE five

Fig. 9 schematically shows a hardware architecture diagram of a computer device suitable for implementing the advertisement recall method according to an embodiment of the present application. In the present embodiment, the computer device 20 is a device capable of automatically performing numerical calculation and/or information processing in accordance with a command set or stored in advance. For example, the server may be a rack server, a blade server, a tower server, or a rack server (including an independent server or a server cluster composed of a plurality of servers). As shown in fig. 9, the computer device 20 includes at least, but is not limited to: the memory 910, processor 920, and network interface 930 may be communicatively linked to each other via a system bus. Wherein:

the memory 910 includes at least one type of computer-readable storage medium including a flash memory, a hard disk, a multimedia card, a card-type memory (e.g., SD or DX memory, etc.), a Random Access Memory (RAM), a Static Random Access Memory (SRAM), a read-only memory (ROM), an electrically erasable programmable read-only memory (EEPROM), a programmable read-only memory (PROM), a magnetic memory, a magnetic disk, an optical disk, etc. In some embodiments, the storage 910 may be an internal storage module of the computer device 20, such as a hard disk or a memory of the computer device 20. In other embodiments, the memory 910 may also be an external storage device of the computer device 20, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card), etc. provided on the computer device 20. Of course, the memory 910 may also include both internal and external memory modules of the computer device 20. In this embodiment, the memory 910 is generally used for storing the operating system and various application software installed in the computer apparatus 20, such as program codes of the advertisement recall method. In addition, the memory 910 may also be used to temporarily store various types of data that have been output or are to be output.

Processor 920 may be, in some embodiments, a Central Processing Unit (CPU), a controller, a microcontroller, a microprocessor, or other data Processing chip. The processor 920 is generally configured to control the overall operation of the computer device 20, such as performing control and processing related to data interaction or communication with the computer device 20. In this embodiment, the processor 920 is configured to execute program codes stored in the memory 910 or process data.

Network interface 930 may include a wireless network interface or a wired network interface, with network interface 930 typically being used to establish communication links between computer device 20 and other computer devices. For example, the network interface 930 is used to connect the computer device 20 to an external terminal via a network, establish a data transmission channel and a communication link between the computer device 20 and the external terminal, and the like. The network may be a wireless or wired network such as an Intranet (Intranet), the Internet (Internet), a Global System of Mobile communication (GSM), Wideband Code Division Multiple Access (WCDMA), a 4G network, a 5G network, Bluetooth (Bluetooth), or Wi-Fi.

It is noted that FIG. 9 only shows a computer device having components 910 and 930, but it is to be understood that not all of the shown components are required and that more or fewer components may be implemented instead.

In this embodiment, the advertisement recall method stored in the memory 910 may be further divided into one or more program modules and executed by one or more processors (in this embodiment, the processor 920) to complete the present invention.

EXAMPLE six

The present embodiments also provide a computer-readable storage medium having stored thereon a computer program which, when executed by a processor, implements the steps of the advertisement recall method in the embodiments.

In this embodiment, the computer-readable storage medium includes a flash memory, a hard disk, a multimedia card, a card type memory (e.g., SD or DX memory, etc.), a Random Access Memory (RAM), a Static Random Access Memory (SRAM), a Read Only Memory (ROM), an Electrically Erasable Programmable Read Only Memory (EEPROM), a Programmable Read Only Memory (PROM), a magnetic memory, a magnetic disk, an optical disk, and the like. In some embodiments, the computer readable storage medium may be an internal storage unit of the computer device, such as a hard disk or a memory of the computer device. In other embodiments, the computer readable storage medium may be an external storage device of the computer device, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card), and the like provided on the computer device. Of course, the computer-readable storage medium may also include both internal and external storage devices of the computer device. In this embodiment, the computer-readable storage medium is generally used for storing an operating system and various types of application software installed in the computer device, for example, the program codes of the advertisement recall method in the embodiment, and the like. Further, the computer-readable storage medium may also be used to temporarily store various types of data that have been output or are to be output.

It will be apparent to those skilled in the art that the modules or steps of the embodiments of the present invention described above may be implemented by a general purpose computing device, may be integrated into a single computing device or distributed across a network of multiple computing devices, and may alternatively be implemented by program code executable by a computing device, such that the steps shown or described may be executed by a computing device stored in a storage device and, in some cases, may be executed out of order from that shown or described, or separately fabricated into individual circuit modules, or fabricated into a single circuit module from multiple modules or steps of the same. Thus, embodiments of the invention are not limited to any specific combination of hardware and software.

The above description is only a preferred embodiment of the present invention, and not intended to limit the scope of the present invention, and all modifications of equivalent structures and equivalent processes, which are made by using the contents of the present specification and the accompanying drawings, or directly or indirectly applied to other related technical fields, are included in the scope of the present invention.