CN113793161A

CN113793161A - Advertisement delivery method, advertisement delivery device, readable storage medium and electronic device

Info

Publication number: CN113793161A
Application number: CN202010646177.4A
Authority: CN
Inventors: 李勇; 赵梓皓; 彭长平; 方治炜; 包勇军; 颜伟鹏
Original assignee: Beijing Wodong Tianjun Information Technology Co Ltd
Current assignee: Beijing Wodong Tianjun Information Technology Co Ltd
Priority date: 2020-07-07
Filing date: 2020-07-07
Publication date: 2021-12-14

Abstract

The disclosure relates to the technical field of computers, and provides an advertisement putting method and device, a computer readable storage medium and an electronic device. Wherein, the method comprises the following steps: responding to the access operation of the advertisement platform, and acquiring a set of tags of a current user; acquiring an advertisement set relevant to a current user, and inquiring a targeted crowd label corresponding to each advertisement in the advertisement set relevant to the current user; and determining that the advertisement of which the targeted crowd label is matched with at least one label in the label set of the current user is a target advertisement, and delivering the target advertisement to the current user. The scheme can improve the accuracy of advertisement putting based on the acquired advertisement set related to the current user.

Description

Advertisement delivery method, advertisement delivery device, readable storage medium and electronic device

Technical Field

The present disclosure relates to the field of computer technologies, and in particular, to an advertisement delivery method, an advertisement delivery device, a computer-readable storage medium, and an electronic device.

Background

Advertisement is one of the important change ways of internet companies, and advertisement delivery can achieve the mutual win and profit effect between an advertisement platform and an advertiser, so that the advertisement has gradually formed an important industry.

Taking a commodity advertisement in an e-commerce platform as an example, in the related art, firstly, a seller constructs an advertisement unit on the e-commerce platform, selects to put a target object, such as a commodity sku (stock keeping unit) in the e-commerce platform, and defines a specific crowd label to determine a crowd which the commodity advertisement is expected to reach; when the current user accesses the E-commerce platform, the E-commerce platform acquires the label set of the current user and retrieves all advertisement units which define the label of the current user so as to put advertisements

When there are too many advertisement units retrieved based on the user tags, limited by system computing resources and fast response, the number of advertisement units returned per user tag and the number of advertisement units returned by all tags need to be limited. The related art is generally to implement a limit on the number according to the bid of each ad unit, i.e., to return the ad unit with the bid high to satisfy the limit on the number.

However, the advertisement delivery method for retrieving advertisements according to the user tags and limiting the quantity according to the advertisement bids has low delivery accuracy, which reduces the click rate of the current user on each advertisement commodity and affects the advertisement delivery effect.

It is to be noted that the information disclosed in the above background section is only for enhancement of understanding of the background of the present disclosure, and thus may include information that does not constitute prior art known to those of ordinary skill in the art.

Disclosure of Invention

The present disclosure is directed to an advertisement delivery method and apparatus, a computer-readable storage medium, and an electronic device, so as to at least improve the problem of low accuracy of advertisement delivery to a certain extent.

Additional features and advantages of the disclosure will be set forth in the detailed description which follows, or in part will be obvious from the description, or may be learned by practice of the disclosure.

According to a first aspect of the present disclosure, there is provided an advertisement delivery method, including:

responding to the access operation of the advertisement platform, and acquiring a set of tags of a current user;

acquiring an advertisement set relevant to a current user, and inquiring a targeted crowd label corresponding to each advertisement in the advertisement set relevant to the current user;

and determining that the advertisement of which the targeted crowd label is matched with at least one label in the label set of the current user is a target advertisement, and delivering the target advertisement to the current user.

In an exemplary embodiment of the present disclosure, based on the foregoing solution, the obtaining of the advertisement set related to the current user includes:

obtaining historical behavior data of a current user, and determining an advertisement set corresponding to the historical behavior data of the current user through an established prediction model of the advertisement related to the user so as to obtain the advertisement set related to the current user.

In an exemplary embodiment of the present disclosure, based on the foregoing scheme, the advertisement includes a commodity advertisement;

the prediction model of the advertisement related to the user comprises at least one of a first index relation model of the commodity and the co-occurrence commodity corresponding to the commodity, a second index relation model of the query word and the co-occurrence commodity corresponding to the query word, and a commodity vector space model.

In an exemplary embodiment of the present disclosure, based on the foregoing scheme, the first index relationship model of the commodity and the co-occurrence commodity corresponding to the commodity is established by:

acquiring a commodity sequence in historical behavior data of each user, and determining co-occurrence commodities of each commodity in the commodity sequence, wherein the interval between the behavior occurrence time of the co-occurrence commodities of each commodity and the behavior occurrence time of each commodity is within a first preset threshold;

traversing the commodity sequence in the historical behavior data of each user to count the co-occurrence frequency of each commodity and each co-occurrence commodity corresponding to each commodity;

and acquiring the co-occurrence commodities with the co-occurrence frequency larger than a second preset threshold value to determine target co-occurrence commodities of the commodities, and establishing a first index relation model of each commodity and each corresponding target co-occurrence commodity.

In an exemplary embodiment of the present disclosure, based on the foregoing solution, the obtaining of the advertisement set related to the current user includes obtaining a commercial advertisement set related to the current user;

the acquiring of the commercial product advertisement set related to the current user comprises:

acquiring a commodity sequence in historical behavior data of a current user, and determining a timestamp of behavior occurrence time of each commodity in the commodity sequence;

and determining the co-occurrence commodities corresponding to the commodities of the timestamp within the preset time according to the first index relation model so as to obtain a commodity advertisement set related to the current user.

In an exemplary embodiment of the present disclosure, based on the foregoing scheme, the second index relationship model of the query term and the co-occurrence product corresponding to the query term is established in the following manner:

acquiring a sequence of query terms searched in historical behavior data of each user and subsequent behavior commodities corresponding to the query terms, and determining co-occurrence commodities of each query term in the sequence of the query terms and the corresponding subsequent behavior commodities, wherein the behavior occurrence time of the co-occurrence commodities of each query term and the behavior occurrence time of the query terms corresponding to the co-occurrence commodities are within a third preset threshold;

traversing the sequence of each query term and the subsequent behavior commodity corresponding to the query term to count the co-occurrence frequency of each query term and each co-occurrence commodity corresponding to the query term;

acquiring the co-occurrence commodities with the co-occurrence frequency larger than a fourth preset threshold value to determine target co-occurrence commodities of the query terms, and establishing a second index relation model of each query term and each target co-occurrence commodity corresponding to the query term.

acquiring a historical query word sequence in historical behavior data of a current user, and determining a timestamp when a search behavior of each query word in the historical query series occurs;

and determining the co-occurrence commodities corresponding to the query terms of the timestamp within the preset time according to the second index relation model so as to obtain a commodity advertisement set related to the current user.

In an exemplary embodiment of the present disclosure, based on the foregoing scheme, the commodity vector space model is established by:

and acquiring commodity sequences in historical behavior data of a user, taking each commodity sequence as training data, and training the machine learning model to obtain a commodity vector space model.

acquiring a commodity sequence in historical behavior data of a current user, and inputting the commodity sequence into the commodity vector space model to obtain commodity vector representation of the current user;

determining similarity between the commodity vector representation of the current user and the vector representation of the commodity in each commodity advertisement;

and acquiring the commodity advertisement corresponding to the commodity with the similarity larger than a fifth preset threshold value so as to determine the commodity advertisement set related to the current user.

In an exemplary embodiment of the present disclosure, based on the foregoing solution, delivering the targeted advertisement to the current user includes:

acquiring an intersection of the targeted population label of the target advertisement and the label in the set of the labels of the current user, and determining the price corresponding to the label with the highest advertisement price in the intersection as the target price of the target advertisement;

predicting the probability of the current user performing behavior operation on each target advertisement, determining the sequence of the display positions of the target advertisements according to the probability and the target price, and delivering the target advertisements to the current user according to the sequence;

wherein the behavior operation comprises at least one of clicking, browsing and purchasing.

According to a second aspect of the present disclosure, there is provided an advertisement delivery apparatus including:

the system comprises a current user tag acquisition module, a current user tag acquisition module and a tag matching module, wherein the current user tag acquisition module is configured to respond to access operation of an advertisement platform and acquire a set of tags of a current user;

the system comprises a targeted crowd tag query module, a target crowd tag query module and a target crowd tag query module, wherein the targeted crowd tag query module is configured to acquire an advertisement set relevant to a current user and query a targeted crowd tag corresponding to each advertisement in the advertisement set relevant to the current user;

and the target advertisement putting module is configured to determine that the advertisement of which the targeted crowd label is matched with at least one label in the current user label set is a target advertisement, and put the target advertisement to the current user.

According to a third aspect of the present disclosure, there is provided a computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements the advertisement delivery method according to the first aspect of the embodiments described above.

According to a fourth aspect of the embodiments of the present disclosure, there is provided an electronic apparatus including: a processor; and a storage device for storing one or more programs which, when executed by the one or more processors, cause the one or more processors to implement the advertisement delivery method as described in the first aspect of the embodiments above.

As can be seen from the foregoing technical solutions, the advertisement delivery method, the advertisement delivery apparatus, and the computer-readable storage medium and the electronic device for implementing the advertisement delivery method in the exemplary embodiments of the present disclosure have at least the following advantages and positive effects:

in the technical solutions provided by some embodiments of the present disclosure, first, in response to an access operation to an advertisement platform, a set of tags of a current user may be obtained, and at the same time, an advertisement set related to the current user may be obtained; then, the targeted crowd label corresponding to each advertisement in the advertisement set relevant to the current user can be inquired; finally, an advertisement whose targeted demographic tag matches at least one tag in the set of tags of the current user may be determined as a targeted advertisement, which is delivered to the current user. Compared with the prior art, on one hand, the method and the device can improve the accuracy of advertisement putting based on the acquired advertisement set related to the current user, and further improve the conversion rate of advertisement putting; on the other hand, the matching of the targeted crowd labels corresponding to the advertisements related to the user and the labels in the label set of the current user can be carried out, so that the efficiency of label matching can be improved, the response speed of the advertisement platform is further improved, and the advertisement platform can carry out advertisement display more quickly.

It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure.

Drawings

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present disclosure and together with the description, serve to explain the principles of the disclosure. It is to be understood that the drawings in the following description are merely exemplary of the disclosure, and that other drawings may be derived from those drawings by one of ordinary skill in the art without the exercise of inventive faculty.

FIG. 1 illustrates a flow diagram of a prior art method of advertisement placement in an exemplary embodiment of the present disclosure;

FIG. 2 illustrates a flow diagram of a method of advertisement delivery in an exemplary embodiment of the present disclosure;

FIG. 3 illustrates a flow diagram of a method of building a first index relationship model in an exemplary embodiment of the present disclosure;

FIG. 4 illustrates a flowchart of a method of building a second index relationship model in an exemplary embodiment of the present disclosure;

FIG. 5 is a flow diagram illustrating a method for obtaining a set of advertisements relevant to a current user according to a commodity vector space model in an exemplary embodiment of the present disclosure;

FIG. 6 shows a schematic structural diagram of an advertisement delivery device in an exemplary embodiment of the present disclosure;

FIG. 7 shows a schematic diagram of a structure of a computer storage medium in an exemplary embodiment of the disclosure; and the number of the first and second groups,

fig. 8 shows a schematic structural diagram of an electronic device in an exemplary embodiment of the present disclosure.

Detailed Description

Example embodiments will now be described more fully with reference to the accompanying drawings. Example embodiments may, however, be embodied in many different forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the concept of example embodiments to those skilled in the art. The described features, structures, or characteristics may be combined in any suitable manner in one or more embodiments. In the following description, numerous specific details are provided to give a thorough understanding of embodiments of the disclosure. One skilled in the relevant art will recognize, however, that the subject matter of the present disclosure can be practiced without one or more of the specific details, or with other methods, components, devices, steps, and the like. In other instances, well-known technical solutions have not been shown or described in detail to avoid obscuring aspects of the present disclosure.

The terms "a," "an," "the," and "said" are used in this specification to denote the presence of one or more elements/components/parts/etc.; the terms "comprising" and "having" are intended to be inclusive and mean that there may be additional elements/components/etc. other than the listed elements/components/etc.; the terms "first" and "second", etc. are used merely as labels, and are not limiting on the number of their objects.

Furthermore, the drawings are merely schematic illustrations of the present disclosure and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and thus their repetitive description will be omitted. Some of the block diagrams shown in the figures are functional entities and do not necessarily correspond to physically or logically separate entities.

Advertisement delivery has gradually formed an important industry as one of the important emerging ways of internet companies. Different businesses in the advertising industry assume different responsibilities and are located in different locations in the advertising industry chain. From the overall marketing market, the industry chain of advertisement placement may include three parts, respectively: an advertiser, an advertisement platform, and a user.

The advertiser hopes that the advertisements put by the advertiser can be pushed to specific crowds for achieving the purpose of marketing, so that the advertiser can select specific crowd labels for products of the advertiser, and the process of selecting the specific crowd labels by the advertiser is called crowd targeting. In order to realize that an advertiser can select a specific crowd label, the advertising Platform completes crowd label processing through a Data Management Platform (DMP) and provides the advertiser with the specific crowd label selection. When a user accesses the advertisement platform, the advertisement platform can mark a proper label for the user according to the behavior and the registration information of the user to generate a user label portrait, and when the user label portrait accords with a specific label defined by an advertiser, the advertisement of the advertiser can be displayed to the user.

After the advertisement putting is completed, the advertisement platform can calculate the profit according to the click rate, and the advertisement putting accuracy is an important influence factor of the click rate.

In the prior art, the accurate advertisement delivery is mainly realized in the modes of enriching user tags and the like, for example, the enriched user tags are constructed, and wider choices and more refined orientation are provided for advertisers; a tool for delineating specific crowds is provided for merchants, and advertisers can conveniently and directionally mine the crowds; according to the users defined by the advertiser, diffusion is completed according to the characteristics of the user tags, and population expansion is completed, so that the advertiser can deliver the advertisements to more target populations.

Fig. 1 shows a flow chart of an advertisement delivery method in the prior art in an exemplary embodiment of the present disclosure. Referring to fig. 1, the method includes steps S110 to S150.

In step S110, the user accesses an advertising platform. Before the user accesses the advertisement platform, the advertiser may construct an advertisement unit on the advertisement platform, select a target object to be delivered, for example, a commodity sku (stock serving unit) in the e-commerce platform, and select a specific crowd label for the target object to be delivered from the crowd labels provided by the advertisement platform to complete the crowd targeting. At the same time, the advertiser may be a different label L_iProviding different advertising bids by corresponding crowds

After the user accesses the advertisement platform, the advertisement platform may obtain a set of user tags in step S120.

In step S130, the advertisement units are retrieved according to the user tag set to obtain an advertisement set.

Wherein, the advertisement platform is according to user label L_iBy indexing<L_i,groupid_1,groupid_2,...,groupid_k>Search advertisement unit, wherein the main key of the index is user label L_iThe value list group _1, group _2, group _ k is the selected label L_iAll of the ad units of (1). When the user has a plurality of labels, all the label retrieval advertisement units of the user are combined to obtain an advertisement unit set T.

In step S140, click through rate estimation is performed on the advertisements in the advertisement set. Specifically, according to the advertisement unit set T, the corresponding delivered object is obtained, and the click probability of the user on the delivered object is estimated.

In step S150, expected revenue is calculated according to the click through rate and the advertisement bids, and the advertisements in the advertisement set are displayed in a sorted manner according to the expected revenue.

For the above step S130, when there are too many advertisement units (millions and above), the number of returned advertisement units per tag and the number of advertisement unit sets T of users need to be limited due to the limitation of computing resources and fast response of the system. In the prior art, when the advertisement units are preferred, the advertisement units are generally sorted according to the bids of the advertisement units, and the unit with the highest bid is preferred. The size of the final preferred set of ad units T is typically on the order of thousands.

However, although the richness of the user tags is guaranteed, the method of optimizing the advertisement units according to the bids still reduces the accuracy of advertisement placement, so that the click rate of the finally placed advertisements is low, and the revenue of the advertisement platform and the experience of the user are affected.

In an exemplary embodiment of the present disclosure, an advertisement delivery method is provided that overcomes, at least to some extent, the above-mentioned deficiencies in the prior art related art.

Fig. 2 shows a flowchart of an advertisement delivery method in an exemplary embodiment of the present disclosure, and referring to fig. 2, the method includes:

step S210, responding to the access operation of the advertisement platform, and acquiring a set of the labels of the current user;

step S220, acquiring an advertisement set relevant to the current user, and inquiring a targeted crowd label corresponding to each advertisement in the advertisement set relevant to the current user;

step S230, determining that the advertisement with the targeted crowd label matched with at least one label in the label set of the current user is a target advertisement, and delivering the target advertisement to the current user.

In the technical solution provided by the embodiment shown in fig. 2, in the technical solution provided by some embodiments of the present disclosure, first, in response to an access operation to an advertisement platform, a set of tags of a current user may be obtained, and at the same time, an advertisement set related to the current user may be obtained; then, the targeted crowd label corresponding to each advertisement in the advertisement set relevant to the current user can be inquired; finally, an advertisement whose targeted demographic tag matches at least one tag in the set of tags of the current user may be determined as a targeted advertisement, which is delivered to the current user. Compared with the prior art, on one hand, the method and the device can improve the accuracy of advertisement putting based on the acquired advertisement set related to the current user, and further improve the conversion rate of advertisement putting; on the other hand, the matching of the targeted crowd labels corresponding to the advertisements related to the user and the labels in the label set of the current user can be carried out, so that the efficiency of label matching can be improved, the response speed of the advertisement platform is further improved, and the advertisement platform can carry out advertisement display more quickly.

The following detailed description of the various steps in the example shown in fig. 2:

in step S210, in response to the access operation to the advertisement platform, a set of tags of the current user is obtained.

The advertising platform may include any streaming media platform including, but not limited to, e-commerce platforms, news media platforms, short video websites, and the like. The access operation to the advertisement platform may include any operation that a user opens the advertisement platform, and the user clicks, browses, views, and the like on the advertisement platform.

In an exemplary embodiment, the tags of the current user include at least one of static attribute tags, behavior class tags, and algorithm mining class tags.

Taking the advertisement platform as an e-commerce platform as an example, the static attribute tags may include gender, age, purchasing power, etc., the behavior class tags may include tags of purchased goods of a specific category, and the algorithm mining class tags may include various tags obtained through an algorithm, such as promotion sensitive population tags, etc.

For example, a specific implementation manner of step S210 may be that, when the user accesses the advertisement platform, the advertisement platform determines the tag of the current user according to the historical behavior information and the registration information of the current user to obtain the set of tags of the current user. Wherein, the set of the current user's tags includes at least one user tag.

For example, the behavior class tags and the algorithm mining class tags may be determined according to historical behavior information of the user, and the static attribute tags of the user are determined according to registration information of the user, so as to generate and obtain a set of tags of the current user.

Specifically, for example, when a user purchases clothing, the behavior class tag of the user may comprise a clothing class tag, and when the user purchases using a coupon, the algorithm mining class tag may comprise a promotion sensitive crowd tag. When the user registers, the user can select a preset number of preset labels by himself or automatically determine the static attribute labels of the user according to the registration information of the user, such as age, gender and the like, so that the set formed by the labels is determined as the set of the labels of the current user.

After the set of tags of the current user is obtained, in step S220, an advertisement set related to the current user is obtained, and a targeted population tag corresponding to each advertisement in the advertisement set related to the current user is queried.

In an exemplary embodiment, the set of advertisements relevant to the current user includes a set of advertisements of interest to the current user as determined by a machine learning algorithm, such as a collaborative filtering algorithm and/or a deep learning model. Of course, it is within the scope of the present disclosure that the advertisement related to the current user may be obtained by any other method.

For example, a specific embodiment of obtaining the advertisement set related to the current user may be to obtain historical behavior data of each user, and establish a prediction model of the advertisement related to the user according to the historical behavior data of each user; obtaining historical behavior data of a current user, and determining an advertisement set corresponding to the historical behavior data of the current user through the established prediction model of the advertisements relevant to the user so as to obtain the advertisement set relevant to the current user.

Wherein the set of advertisements relevant to the current user may include a set of commercial advertisements relevant to the current user. Further, the prediction model of the advertisement related to the user may include at least one of a first index relation model of the product and the co-occurring product corresponding thereto, a second index relation model of the query term and the co-occurring product corresponding thereto, and a product vector space model.

For example, fig. 3 is a flowchart illustrating a method for building a first index relationship model according to an exemplary embodiment of the disclosure. Referring to fig. 3, the method may include steps S310 to S330.

In step S310, a product sequence in the historical behavior data of each user is acquired, and a co-occurrence product of each product in the product sequence is determined.

And the interval between the action occurrence time of the co-occurrence commodity of each commodity and the action occurrence time of each commodity is within a first preset threshold value. In other words, the co-occurrence merchandise of each merchandise may include the merchandise whose interval between the action occurrence time and the action occurrence time of the merchandise is within the first preset threshold.

In an exemplary embodiment, the commodity sequence in the historical behavior data of each user may be a commodity sequence in the historical behavior data of all users in the advertisement platform, may also be a commodity sequence in the historical behavior data of all users with special authority, and may also be a commodity sequence in the historical behavior data of some users in the preset advertisement platform. The present exemplary embodiment is not particularly limited in this regard.

The commodity sequence in the historical behavior data of the user may be a commodity sequence generated by any historical triggering operation on commodities, such as clicking, browsing, buying, purchasing and the like of the user.

For example, the historical behaviors of different users may generate different commodity sequences, and taking the example that the advertisement platform is an e-commerce platform, each user is all users of the e-commerce platform, and the historical behavior of the user is a click commodity, the specific implementation of step S310 may be to obtain a click commodity sequence in the historical behavior data of all users of the e-commerce platform, and determine a co-occurrence commodity of each commodity in all commodity sequences.

Wherein each commodity may include any commodity in all of the sequences of commodities. Each commodity and its corresponding co-occurrence commodity may be commodities whose timestamps are different when two commodity behaviors occur in each commodity sequence. Generally, the timestamp of each item is less than the co-occurring item for each item.

For example, the commodity click sequence corresponding to the historical behavior data of the user 1 may be "item 1, time2, item3, …, itemi, itemN". If the difference between the time stamps of the click actions of item1 and time2 occurs within a first preset threshold, such as 600 seconds, item1 and item2 co-occur, and item2 can be considered a co-occurring item of item 1. Of course, if the difference between the time stamps at which the click behaviors of item1 and item3 occur is within the first preset threshold, item3 can be considered as another co-occurring item of item1 at this time.

Additionally, item1 may also co-occur with other items in an item sequence of historical behavioral data of other users. That is, there may be a plurality of co-occurring products for each product, and the co-occurring product for each product may be determined based on the product sequence in the historical behavior data of all users.

After the co-occurrence of each commodity is determined, in step S320, the commodity sequence in the historical behavior data of each user is traversed to count the co-occurrence frequency of each commodity and each co-occurrence commodity corresponding to each commodity.

Here, the product sequence in the historical behavior data of each user is already described in step S310, and is not described here again.

For example, the specific implementation of step S320 may be that the commodity sequence in the historical behavior data of all the users is traversed to count the occurrence frequency of each co-occurring commodity corresponding to each commodity with the co-occurring commodity, and the co-occurring frequency is determined according to the occurrence frequency.

For example, item1 and item2 in the commodity sequence corresponding to the historical behavior data of the first user co-occur, item2 is the co-occurring commodity of item1, and the 1 st occurrence, then the co-occurrence frequency of item1 and item2 may be 1 at this time. When item1 and item2 co-occur in the commodity sequence of the second user's historical behavior data, item2 is co-occurring commodity of item1, and the co-occurrence frequency of item1 and item2 is increased by 1 correspondingly at the 2 nd occurrence. By analogy, when there are N users, traversing the commodity sequence in the historical behavior data of the N users, and counting the number of occurrences of item2 when item2 in the historical behavior data of the N users is taken as a co-occurring commodity of item1, it is possible to determine the co-occurring frequency of item1 and its corresponding co-occurring commodity item 2. The co-occurrence frequency of other commodities and the commodities corresponding to the other commodities is the same as the determination method of the co-occurrence frequency of item1 and the co-occurrence commodity item2 corresponding to the other commodities, and the details are not repeated here.

After the co-occurrence frequency of each commodity and each co-occurrence commodity corresponding to each commodity is counted, in step S330, co-occurrence commodities of which the co-occurrence frequency is greater than a second preset threshold are obtained to determine a target co-occurrence commodity of each commodity, and a first index relationship model of each commodity and each target co-occurrence commodity corresponding to each commodity is established.

The first index relationship model may include a model of an index relationship between each commodity and each target co-occurrence commodity corresponding to the commodity.

For example, in the above-mentioned specific implementation of step S330, a co-occurring commodity with a co-occurring frequency greater than a second preset threshold is selected from all co-occurring commodities in each commodity, and is used as a target co-occurring commodity of the commodity, each commodity is used as a main key, the target co-occurring commodity of each commodity is used as a value list, and an index relationship model between each commodity and each corresponding target co-occurring commodity is established.

For example, the specific implementation manner of determining the target co-occurrence product may also be that the co-occurrence frequencies of the product and the corresponding co-occurrence products are sorted in descending order, and N co-occurrence products with the co-occurrence frequencies sorted in the top N are determined as the target co-occurrence product. Of course, the co-occurrence product with a high co-occurrence frequency may be selected from the co-occurrence products as the target co-occurrence product in any other manner, and this is not particularly limited in the present exemplary embodiment.

Generally, the co-occurrence frequency can represent the co-occurrence similarity between a commodity and its corresponding co-occurrence commodity, and the larger the value of the co-occurrence frequency is, the higher the co-occurrence similarity is. The higher the co-occurrence similarity is, the higher the correlation between the co-occurrence product and each product in the product sequence in the historical behavior data of the user is, the higher the probability that the user clicks, browses, purchases, and purchases the product is, and the higher the accuracy of the product advertisement placement is. Therefore, the accuracy of advertisement delivery can be improved by screening the co-occurrence frequency.

In addition, due to the fact that co-occurrence frequency screening is conducted, the number of advertisements can be limited, and therefore the advertisement putting is selected in a relatively small number, accuracy can be improved, and meanwhile the response speed of the system is improved.

For example, after the first index relationship model of each product and each target co-occurring product corresponding to each product is established through the above steps S310 to S330, the product advertisement set related to the current user may be obtained through the above first index relationship model.

Specifically, an implementation manner of acquiring a commercial advertisement set related to a current user may be that, first, a commercial sequence in historical behavior data of the current user is acquired, and a timestamp of a behavior occurrence time of each commercial in the commercial sequence is determined; then, according to the first index relation model, co-occurring commodities corresponding to commodities with time stamps within preset time are determined, so that a commodity advertisement set relevant to the current user is obtained.

In an exemplary embodiment, the commodity sequence in the historical behavior data of the current user may include a commodity sequence generated by any historical triggering operation on commodities, such as clicking, browsing, buying, purchasing and the like, of the current user.

Each commodity in the commodity sequence has a timestamp of a corresponding action occurrence time, for example, if commodity 1 in the commodity sequence is clicked at time T1, the timestamp of the action occurrence time corresponding to commodity 1 is T1; when the product 2 in the product behavior sequence is purchased at time T2, the time stamp of the behavior occurrence time of the product 2 is T2, or the like.

Generally, the closer the behavior occurrence timestamp and the current timestamp (which may be the timestamp of when the user accessed the advertising platform) in the sequence of items of historical behavior data of the current user are, the higher the relevance to the current user is.

Therefore, when the product advertisement set related to the current user is obtained, according to the first index relationship model, the co-occurring product corresponding to each product with the timestamp within the preset time can be determined, so as to obtain the product advertisement set related to the current user. Wherein, each commodity in the preset time can be each commodity in the last N days.

For example, an embodiment of obtaining a commercial advertisement set related to a current user may also be that, first, a commercial sequence in historical behavior data of the current user is obtained, and a timestamp of a behavior occurrence time of each commercial in the commercial sequence is determined; then, a current timestamp of a current user when accessing the advertisement platform is obtained, a difference value between the current timestamp and the timestamp of the behavior occurrence time of each commodity is determined, the commodities in the commodity sequence of the historical behavior data of the current user are sequenced according to the sequence from small to large of the difference value, the co-occurring commodities of the sequenced commodities are sequentially determined through the first index relation model until the quantity of the commodities in the obtained commodity set reaches a preset quantity, and therefore the commodity advertisement set related to the user is obtained.

For example, the preset number may be 300, and the commodity sequence obtained by sorting the timestamp differences from small to large is "item 3, item4, item2, item7, item …, item" so that 20 co-occurring commodities of item3 can be obtained first through the first index relationship, and if the preset number 300 is not reached, the co-occurring commodities of item4, item2, and item7 are continuously obtained in sequence until the number of the obtained commodity advertisements reaches 300.

However, when co-occurrence commodities of the commodities are acquired, the co-occurrence commodities can be acquired in descending order of co-occurrence frequency, for example, after all target co-occurrence commodities of item2 are acquired, the number of the current commodity advertisement set is 270, and when the target co-occurrence commodity of item7 is greater than 30, the first 30 target co-occurrence commodities are acquired.

Further, fig. 4 is a flowchart illustrating a method for establishing a second index relationship model in an exemplary embodiment of the disclosure. Referring to fig. 4, the method may include steps S410 to S430.

In step S410, a sequence of the query term searched in the historical behavior data of each user and the subsequent behavior commodity corresponding thereto is obtained, and the co-occurrence commodity of each query term is determined in the sequence of the query term and the subsequent behavior commodity corresponding thereto.

And the action occurrence time of the co-occurrence commodity of each query term and the action occurrence time of the corresponding query term are within a third preset threshold value. The third preset threshold may be equal to or different from the first preset threshold, and this is not particularly limited in this exemplary embodiment.

In an exemplary embodiment, the subsequent behavior commodity corresponding to the query term may include a commodity corresponding to a behavior operation such as clicking, browsing, purchasing, and purchasing by the user after the query term is searched.

The form of the sequence of the query term and the corresponding follow-up behavior commodity may be "query, item1, item2, …, itemN", where the query is the query term, and item1, item2, …, itemN are the corresponding follow-up behavior commodity of the query term.

In an exemplary embodiment, the sequence of the query term and the subsequent behavioral good corresponding to the query term searched in the historical behavioral data of each user may be a sequence of the query term and the subsequent behavioral good corresponding to the query term in the historical behavioral data of all users in the advertisement platform, a sequence of the query term and the subsequent behavioral good corresponding to the query term in the historical behavioral data of all users with special authority, or a sequence of the query term and the subsequent behavioral good corresponding to the query term in the historical behavioral data of a part of users in a preset advertisement platform.

The sequence of the query terms and the corresponding subsequent behavior commodities in the historical data of each user can be multiple. The sequence of the query term in the historical behavior data of each user and the subsequent behavior commodity corresponding to the query term can also be the sequence of the query term meeting the preset condition in the historical behavior data of all users or part of users and the subsequent behavior commodity corresponding to the query term. For example, the query term in the clothing category and the corresponding subsequent behavior commodity sequence of the query term. The present exemplary embodiment is not particularly limited in this regard.

For example, a specific implementation manner of determining the co-occurrence product of each query term in the sequence of the query term and the corresponding subsequent behavior product may be to obtain a sequence of all query terms and corresponding subsequent behavior products in the history data of all or part of the users in the advertisement platform, and determine the co-occurrence product of each query term in the sequence of all or part of the query terms and corresponding subsequent behavior products of all or part of the users.

The subsequent behavior commodities corresponding to each query term in different user sequences can be different, and the number of the co-occurrence commodities corresponding to each query term can be multiple.

After the co-occurrence commodities of the query terms are determined, in step S420, the sequence of the query terms and the subsequent behavior commodities corresponding to the query terms is traversed to count the co-occurrence frequency of the query terms and the co-occurrence commodities corresponding to the query terms.

For example, the specific implementation manner of step S420 may be to traverse all query terms in the historical behavior data of all users and the sequence of subsequent behavior commodities corresponding to the query terms, to count the occurrence frequency of each co-occurrence commodity corresponding to each query term with the co-occurrence commodity, and determine the co-occurrence frequency according to the occurrence frequency.

For example, the sequence of the query word and its corresponding subsequent behavior commodity in the historical behavior data of the first user may be 2, which are "query 1, item1, item2, …, item" and "query 2, item4, item7, …, and item", respectively, in the first sequence, commodity item1 and query term query1 co-occur, item1 is a co-occurring commodity of query1, and the 1 st occurrence, then the co-occurrence frequency of item1 and query1 is 1 at this time; in the second sequence, item4 and query2 co-occur, item4 co-occurring as a co-occurring commodity of query2, at 1 st occurrence, then the co-occurrence frequency of item4 and query2 is now 1. When item1 and query1 are also present in the historical behavior data of the second user, at this time, item1 is taken as a co-occurrence commodity of query1, and the 2 nd occurrence is carried out, the co-occurrence frequency of item1 and query1 is correspondingly increased by 1, when item4 and query2 are not present in the historical behavior data of the second user, the co-occurrence frequency of item4 and query2 is unchanged, and the corresponding co-occurrence frequency is increased by 1 until item4 and query2 are present in other sequences. Of course, new query terms and their corresponding co-occurring products may also appear in the second user's historical behavior data. By analogy, when there are N users, traversing the commodity sequence in the historical behavior data of the N users, and counting the number of occurrences of item1 when item1 in the historical behavior data of the N users is used as a co-occurring commodity of query1, it is possible to determine the co-occurring frequency of query1 and its corresponding co-occurring commodity item 1. The co-occurrence frequency of other query terms and their corresponding co-occurrence items is the same as the determination method of the co-occurrence frequency of query1 and its corresponding co-occurrence item1, and is not described here again.

After the co-occurrence frequency of each query term and each co-occurrence product corresponding to the query term is counted, in step S430, the co-occurrence product with the co-occurrence frequency greater than a fourth preset threshold is obtained to determine the target co-occurrence product of the query term, and a second index relationship model of each query term and each target co-occurrence product corresponding to the query term is established.

The second index relationship model may include a model of an index relationship between each query term and each target co-occurring product corresponding to the query term.

For example, the specific implementation of step S430 is the same as the specific implementation of step S330 described above, and the technical term "each product" in step S330 is replaced by "query word", and "each product and its corresponding co-occurring product" is replaced by "each query word and its corresponding co-occurring product", which is not described herein again.

For example, through the steps S410 to S430 described above, after the first index relationship model of each query term and each target co-occurring product corresponding to the query term is established, the product advertisement set related to the current user may be obtained.

Specifically, the embodiment of obtaining the commercial advertisement set related to the current user may be that, first, a historical query word sequence in the historical behavior data of the current user is obtained, and a timestamp when a search behavior of each query word in the historical query series sequence occurs is determined; and determining the co-occurrence commodities corresponding to the query terms of the timestamp within the preset time according to the second index relation model so as to obtain a commodity advertisement set related to the current user.

Each query word in the historical query word sequence has a timestamp of corresponding search behavior occurrence time, for example, if a query word 1 is searched at time T1, the timestamp of the behavior occurrence time corresponding to the query word 1 is T1; when the query word 2 is searched at time T2, the timestamp of the behavior occurrence time of the query word 2 is T2, and the like.

Generally, query words in the historical query word sequence of the historical behavior data of the current user, which have a behavior occurrence timestamp close to the current timestamp (which may be the timestamp of when the user accessed the advertisement platform), have higher relevance to the current user.

Therefore, when the product advertisement set related to the current user is obtained, according to the second index relationship model, the co-occurring product corresponding to each query term with the timestamp within the preset time may be determined, so as to obtain the product advertisement set related to the current user. Each query term in the preset time can be the query term in the last N days.

Illustratively, when the advertisement set related to the current user is obtained according to the first index relationship model of each query word and each target co-occurrence product corresponding to the query word, if the advertisement set is in a recommendation scene, that is, when the current user does not input any query word, the advertisement set related to the current user can be obtained according to the historical query word sequence of the current user; if the advertisement set is in a search scene, namely when the current user inputs the query word to search, the advertisement set related to the current user can be obtained according to the query word of the current search request.

For example, the embodiment of obtaining the commercial advertisement set related to the current user may further include obtaining a historical query word sequence in the historical behavior data of the current user, and determining a timestamp of a behavior occurrence time of each query word in the historical query word sequence of the merchant; then, a current timestamp of a current user when accessing the advertisement platform is obtained, a difference value between the current timestamp and the timestamp of the behavior occurrence time of each query word is determined, the query words in the historical behavior data of the current user are sequenced according to the sequence of the difference value from small to large, co-occurring commodities of the sequenced query words are sequentially determined through the second index relation model until the quantity of the commodities in the obtained commodity set reaches a preset quantity, and accordingly the commodity advertisement set related to the user is obtained.

FIG. 5 illustrates a method for obtaining a set of commercial advertisements associated with a current user via a vector space model in an exemplary embodiment of the disclosure. Illustratively, referring to FIG. 5, the method may include steps S510-S530.

In step S510, a commodity sequence in the historical behavior data of the current user is obtained, and the commodity sequence is input into the commodity vector space model to obtain a commodity vector representation of the current user.

In an exemplary embodiment, the commodity sequence in the historical behavior data of the current user may be a commodity sequence within a preset time, for example, a commodity sequence in the historical behavior data of the last 50 days. Wherein, the preset time can be set according to the requirement.

The commodity vector space model may be built prior to inputting the commodity sequence into the commodity vector space model.

For example, the specific implementation of establishing the commodity vector space model may be: and acquiring commodity sequences in the historical behavior data of the user, taking each commodity sequence as training data, and training the machine learning model to obtain a commodity vector space model.

In an exemplary embodiment, the machine learning model may include any deep learning model, such as a GRU (Gate current Unit) model or the like.

For example, the sequence of items in the historical behavior data of all users in the advertisement platform may be obtained, which may be any sequence of items clicked, browsed, or purchased by the users. And inputting the commodity sequence of each user as training data into the GRU model so as to perform offline training on the GRU model. Taking the commodity sequence of "item 1, time2, item3, … itemi, itemN" as an example, the commodity sequence can be used as a training data, and the commodity sequences of different users can be used as different training data of the GRU model.

When performing the off-line training, the commodity sequence corresponding to each training data may be vectorized to obtain the commodity vector sequence of each commodity sequence. Each commodity vector sequence is then input into the initial GRU model as training data to enable the GRU model to predict the commodity vector representation for any user from that user's commodity sequence by off-line training.

The size of the commodity vector representation can be set according to actual needs, and for example, the commodity vector representation can be 32-dimensional.

After the commodity vector space model is obtained, the commodity sequence in the historical behavior data of the current user can be obtained, the commodity sequence is vectorized to obtain the commodity vector sequence of the commodity sequence, and then the commodity vector sequence is input into the commodity vector space model to obtain the commodity vector representation of the current user. The commodity vector representation of the current user can be used as the vector representation of the current user.

After the product vector representation of the current user is obtained, in step S520, the similarity between the product vector representation of the current user and the vector representation of the product in each product advertisement is determined.

The similarity in step S520 may be determined according to any method capable of measuring the similarity between two vectors.

For example, the specific implementation manner of step S520 may be to obtain a vector representation of a product in each product advertisement in the advertisement platform, and calculate a cosine similarity between the product vector representation of the current user and the vector representation of the product in each product advertisement.

After the similarity is determined, in step S530, the commercial advertisement corresponding to the commercial product with the similarity greater than the fifth preset threshold is obtained, so as to determine the commercial advertisement set related to the current user. The fifth preset threshold may be set according to actual conditions.

For example, in the specific implementation manner of determining the commercial advertisement set related to the current user in step S530, after the similarity is determined, the commercial advertisements are sorted according to the descending order of the similarity, and N commercial advertisements with top-N similarity ranks are obtained, so as to determine the commercial advertisement set related to the current user.

It should be noted that the exemplary embodiment merely illustrates the method for obtaining the commercial product set related to the current user in each of the methods shown in fig. 3-5, and any other method may be used to achieve the purpose of obtaining the commercial product set related to the current user and still be within the scope of the present disclosure.

Meanwhile, the finally determined commodity advertisement set related to the current user may be a union or an intersection of the commodity advertisement sets related to the current user acquired through a plurality of methods. For example, the union of the advertisement sets related to the current user obtained by the methods shown in fig. 3-5 described above serves as the target advertisement set related to the current user. In this way, the advertisement set relevant to the current user can be obtained from the perspective of multiple correlations, so that the correlation degree of the obtained advertisement set is improved, and the accuracy of advertisement delivery is improved.

After the advertisement set relevant to the current user is obtained, the targeted crowd label corresponding to each advertisement in the advertisement set relevant to the current user can be inquired.

Specifically, the targeted crowd tag corresponding to each advertisement can be queried for the specific crowd tag defined for each advertisement according to the advertiser.

After the targeted crowd tag corresponding to each advertisement is queried, in step S230, an advertisement whose targeted crowd tag matches at least one tag in the set of tags of the current user is determined as a targeted advertisement, and the targeted advertisement is delivered to the current user.

In an exemplary embodiment, the targeted demographic tag for each advertisement may be one or more than one. Different targeted crowd labels can correspond to different advertisement bids and can also correspond to the same advertisement bid.

For example, a specific embodiment of determining a target advertisement may be to obtain an intersection of the targeted demographic tag of each advertisement and a tag in the set of tags of the current user, and determine that the advertisement is the target advertisement when the intersection is not empty.

For example, a specific implementation manner of delivering the target advertisement to the current user may be to obtain an intersection of the targeted population tag of the target advertisement and a tag in the set of tags of the current user, and determine that a price corresponding to a tag with the highest advertisement price in the intersection is the target price of the target advertisement; and predicting the probability of the current user performing behavior operation on each target advertisement, determining the sequence of the display positions of each target advertisement according to the probability and the target price, and delivering each target advertisement to the current user according to the sequence. The behavior operation comprises at least one of clicking, browsing and purchasing.

In general, an advertiser may delineate multiple targeted demographic tags for each advertisement and determine an advertisement price for the multiple targeted demographic tags. In other words, each advertisement corresponds to a plurality of targeted demographic tags, each targeted demographic tag having a corresponding advertisement price. The label with the highest price corresponding to the label of the targeted population in each targeted advertisement can be used as the targeted label of the advertisement, and the price corresponding to the targeted label can be used as the targeted price of the targeted advertisement.

The method comprises the steps of predicting the probability of the behavior operation of a current user on the target advertisement according to the frequency of the historical behavior operation of a tag crowd corresponding to the target tag on the target advertisement, then determining the expected income of each target advertisement according to the probability and the expected value of the target price, sequencing the positions of the target advertisements according to the sequence of the expected income from large to small, and displaying the target advertisements in an advertisement platform interface of the current user according to the position sequencing. Of course, the probability of the behavior operation of the target advertisement by the current user may also be predicted according to other manners, and this exemplary embodiment is not particularly limited in this respect.

Through the steps S210 to S230, the correlation between the current user and the advertisement can be improved according to the obtained advertisement set related to the current user, the problem of poor interpretability of the existing advertisement putting method is effectively solved, and meanwhile, the accuracy of advertisement putting can be improved; in addition, through screening the target advertisements, the user tags and the targeted crowd tags selected by the advertisers can be efficiently matched, and the advertisement putting efficiency is further improved.

Furthermore, through the matching of the user tags and the directional crowd tags, the advertisements can be directionally delivered to the crowd corresponding to the crowd tags defined by the advertisers, the advertisement delivery accuracy is improved, and the user experience can be improved while the demands of the advertisers are met.

Those skilled in the art will appreciate that all or part of the steps implementing the above embodiments are implemented as computer programs executed by a CPU. The computer program, when executed by the CPU, performs the functions defined by the method provided by the present invention. The program may be stored in a computer readable storage medium, which may be a read-only memory, a magnetic or optical disk, or the like.

Furthermore, it should be noted that the above-mentioned figures are only schematic illustrations of the processes involved in the method according to exemplary embodiments of the invention, and are not intended to be limiting. It will be readily understood that the processes shown in the above figures are not intended to indicate or limit the chronological order of the processes. In addition, it is also readily understood that these processes may be performed synchronously or asynchronously, e.g., in multiple modules.

Fig. 6 shows a schematic structural diagram of an advertisement delivery device in an exemplary embodiment of the present disclosure. Referring to fig. 6, the advertisement delivery apparatus 600 may include a current user tag obtaining module 610, a targeted crowd tag querying module 620, and a targeted advertisement delivery module 630. Wherein:

the current user tag obtaining module 610 is configured to respond to an access operation on the advertisement platform to obtain a set of tags of a current user;

the targeted crowd tag query module 620 is configured to obtain an advertisement set related to a current user, and query a targeted crowd tag corresponding to each advertisement in the advertisement set related to the current user;

the targeted advertisement delivery module 630 is configured to determine that the advertisement with the targeted demographic tag matching with at least one tag in the set of tags of the current user is a targeted advertisement, and deliver the targeted advertisement to the current user.

In some embodiments of the present disclosure, based on the foregoing solution, the above-mentioned directional crowd tag query module 620 is further specifically configured to:

acquiring historical behavior data of each user, and establishing a prediction model of advertisements related to the user according to the historical behavior data of each user;

and acquiring historical behavior data of the current user, and determining an advertisement set corresponding to the historical behavior data of the current user through the prediction model of the advertisement related to the user so as to acquire the advertisement set related to the current user.

In some embodiments of the disclosure, based on the foregoing solution, the advertisement includes a product advertisement, and the prediction model of the advertisement related to the user includes at least one of a first index relationship model of a product and a co-occurring product corresponding to the product, a second index relationship model of a query term and a co-occurring product corresponding to the query term, and a product vector space model.

In some embodiments of the present disclosure, based on the foregoing scheme, the first index relationship model of the commodity and the co-occurrence commodity corresponding to the commodity is established by:

In some embodiments of the present disclosure, based on the foregoing solution, the obtaining of the advertisement set related to the current user includes obtaining a commercial advertisement set related to the current user;

In some embodiments of the present disclosure, based on the foregoing scheme, the second index relationship model of the query term and the co-occurrence product corresponding to the query term is established by:

In some exemplary embodiments of the present disclosure, based on the foregoing solution, the obtaining of the advertisement set related to the current user includes obtaining a commercial advertisement set related to the current user;

In some exemplary embodiments of the present disclosure, based on the foregoing scheme, the above commodity vector space model is established by:

In some exemplary embodiments of the present disclosure, based on the foregoing solution, the delivering the targeted advertisement to the current user includes:

The specific details of each unit in the advertisement delivery device have been described in detail in the corresponding advertisement delivery method, and therefore are not described herein again.

It should be noted that although in the above detailed description several modules or units of the device for action execution are mentioned, such a division is not mandatory. Indeed, the features and functionality of two or more modules or units described above may be embodied in one module or unit, according to embodiments of the present disclosure. Conversely, the features and functions of one module or unit described above may be further divided into embodiments by a plurality of modules or units.

Moreover, although the steps of the methods of the present disclosure are depicted in the drawings in a particular order, this does not require or imply that the steps must be performed in this particular order, or that all of the depicted steps must be performed, to achieve desirable results. Additionally or alternatively, certain steps may be omitted, multiple steps combined into one step execution, and/or one step broken down into multiple step executions, etc.

Through the above description of the embodiments, those skilled in the art will readily understand that the exemplary embodiments described herein may be implemented by software, or by software in combination with necessary hardware. Therefore, the technical solution according to the embodiments of the present disclosure may be embodied in the form of a software product, which may be stored in a non-volatile storage medium (which may be a CD-ROM, a usb disk, a removable hard disk, etc.) or on a network, and includes several instructions to enable a computing device (which may be a personal computer, a server, a mobile terminal, or a network device, etc.) to execute the method according to the embodiments of the present disclosure.

In an exemplary embodiment of the present disclosure, there is also provided a computer storage medium capable of implementing the above method. On which a program product capable of implementing the above-described method of the present specification is stored. In some possible embodiments, various aspects of the disclosure may also be implemented in the form of a program product comprising program code for causing a terminal device to perform the steps according to various exemplary embodiments of the disclosure described in the "exemplary methods" section above of this specification, when the program product is run on the terminal device.

Referring to fig. 7, a program product 700 for implementing the above method according to an embodiment of the present disclosure is described, which may employ a portable compact disc read only memory (CD-ROM) and include program code, and may be run on a terminal device, such as a personal computer. However, the program product of the present disclosure is not limited thereto, and in this document, a readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device.

The program product may employ any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the foregoing. More specific examples (a non-exhaustive list) of the readable storage medium include: an electrical connection having one or more wires, a portable disk, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

A computer readable signal medium may include a propagated data signal with readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated data signal may take many forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A readable signal medium may also be any readable medium that is not a readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device.

Program code embodied on a readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.

Program code for carrying out operations for the present disclosure may be written in any combination of one or more programming languages, including an object oriented programming language such as Java, C + + or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computing device, partly on the user's device, as a stand-alone software package, partly on the user's computing device and partly on a remote computing device, or entirely on the remote computing device or server. In the case of a remote computing device, the remote computing device may be connected to the user computing device through any kind of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or may be connected to an external computing device (e.g., through the internet using an internet service provider).

In addition, in an exemplary embodiment of the present disclosure, an electronic device capable of implementing the above method is also provided.

As will be appreciated by one skilled in the art, aspects of the present disclosure may be embodied as a system, method or program product. Accordingly, various aspects of the present disclosure may be embodied in the form of: an entirely hardware embodiment, an entirely software embodiment (including firmware, microcode, etc.) or an embodiment combining hardware and software aspects that may all generally be referred to herein as a "circuit," module "or" system.

An electronic device 800 according to this embodiment of the disclosure is described below with reference to fig. 8. The electronic device 800 shown in fig. 8 is only an example and should not bring any limitations to the functionality and scope of use of the embodiments of the present disclosure.

As shown in fig. 8, electronic device 800 is in the form of a general purpose computing device. The components of the electronic device 800 may include, but are not limited to: the at least one processing unit 810, the at least one memory unit 820, a bus 830 connecting various system components (including the memory unit 820 and the processing unit 810), and a display unit 840.

Wherein the storage unit stores program code that is executable by the processing unit 810 to cause the processing unit 810 to perform steps according to various exemplary embodiments of the present disclosure as described in the "exemplary methods" section above in this specification. For example, the processing unit 810 may perform the following as shown in fig. 2: step S210, responding to the access operation of the advertisement platform, and acquiring a set of the labels of the current user; step S220, acquiring an advertisement set relevant to the current user, and inquiring a targeted crowd label corresponding to each advertisement in the advertisement set relevant to the current user; step S230, determining that the advertisement with the targeted crowd tag matching with at least one tag in the set of tags of the current user is a target advertisement, and delivering the target advertisement to the current user.

As another example, the processing unit 810 may perform various steps as shown in fig. 3, 4, and 5.

The storage unit 820 may include readable media in the form of volatile memory units such as a random access memory unit (RAM)8201 and/or a cache memory unit 8202, and may further include a read only memory unit (ROM) 8203.

The storage unit 820 may also include a program/utility 8204 having a set (at least one) of program modules 8205, such program modules 8205 including, but not limited to: an operating system, one or more application programs, other program modules, and program data, each of which, or some combination thereof, may comprise an implementation of a network environment.

Bus 830 may be any of several types of bus structures including a memory unit bus or memory unit controller, a peripheral bus, an accelerated graphics port, a processing unit, or a local bus using any of a variety of bus architectures.

The electronic device 800 may also communicate with one or more external devices 900 (e.g., keyboard, pointing device, bluetooth device, etc.), with one or more devices that enable a user to interact with the electronic device 800, and/or with any devices (e.g., router, modem, etc.) that enable the electronic device 800 to communicate with one or more other computing devices. Such communication may occur via input/output (I/O) interfaces 850. Also, the electronic device 800 may communicate with one or more networks (e.g., a Local Area Network (LAN), a Wide Area Network (WAN), and/or a public network, such as the internet) via the network adapter 860. As shown, the network adapter 860 communicates with the other modules of the electronic device 800 via the bus 830. It should be appreciated that although not shown, other hardware and/or software modules may be used in conjunction with the electronic device 800, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems, among others.

Through the above description of the embodiments, those skilled in the art will readily understand that the exemplary embodiments described herein may be implemented by software, or by software in combination with necessary hardware. Therefore, the technical solution according to the embodiments of the present disclosure may be embodied in the form of a software product, which may be stored in a non-volatile storage medium (which may be a CD-ROM, a usb disk, a removable hard disk, etc.) or on a network, and includes several instructions to enable a computing device (which may be a personal computer, a server, a terminal device, or a network device, etc.) to execute the method according to the embodiments of the present disclosure.

Furthermore, the above-described figures are merely schematic illustrations of processes included in methods according to exemplary embodiments of the present disclosure, and are not intended to be limiting. It will be readily understood that the processes shown in the above figures are not intended to indicate or limit the chronological order of the processes. In addition, it is also readily understood that these processes may be performed synchronously or asynchronously, e.g., in multiple modules.

Other embodiments of the disclosure will be apparent to those skilled in the art from consideration of the specification and practice of the disclosure disclosed herein. This application is intended to cover any variations, uses, or adaptations of the disclosure following, in general, the principles of the disclosure and including such departures from the present disclosure as come within known or customary practice within the art to which the disclosure pertains. It is intended that the specification and examples be considered as exemplary only, with a true scope and spirit of the disclosure being indicated by the following claims.

Claims

1. An advertisement delivery method, comprising:

2. The method of claim 1, wherein the obtaining a set of advertisements relevant to a current user comprises:

3. The advertisement delivery method according to claim 2, wherein the advertisement comprises a commercial advertisement;

4. The advertisement delivery method according to claim 3, wherein the first index relationship model of the product and its corresponding co-occurrence product is established by:

5. The advertisement delivery method according to claim 4, wherein the obtaining of the advertisement set related to the current user comprises obtaining a commercial advertisement set related to the current user;

6. The advertisement delivery method according to claim 3, wherein the second index relationship model of the query term and the co-occurring product corresponding thereto is established by:

7. The advertisement delivery method according to claim 6, wherein the obtaining of the advertisement set related to the current user comprises obtaining a commercial advertisement set related to the current user;

8. The advertisement delivery method according to claim 3, wherein the commodity vector space model is established by:

9. The method of claim 8, wherein the obtaining a set of advertisements related to a current user comprises obtaining a set of commercial advertisements related to a current user;

10. The advertisement delivery method according to claim 1, wherein delivering the targeted advertisement to the current user comprises:

11. An advertisement delivery device, comprising:

12. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out an advertisement delivery method according to any one of claims 1 to 10.

13. An electronic device, comprising:

one or more processors;

storage means for storing one or more programs which, when executed by the one or more processors, cause the one or more processors to implement the advertisement delivery method of any one of claims 1 to 10.