CN114677169A

CN114677169A - Advertisement creative generation method and device

Info

Publication number: CN114677169A
Application number: CN202210261717.6A
Authority: CN
Inventors: 封战胜
Original assignee: Beijing QIYI Century Science and Technology Co Ltd
Current assignee: Beijing QIYI Century Science and Technology Co Ltd
Priority date: 2022-03-16
Filing date: 2022-03-16
Publication date: 2022-06-28

Abstract

The embodiment of the invention provides an advertisement creative generation method and device, relating to the technical field of data processing, wherein the method comprises the following steps: obtaining a target Uniform Resource Locator (URL) of a webpage of a target product; acquiring the webpage content of the webpage of the target product according to the target URL; extracting target product information of the target product from the webpage content according to creative items required by the advertisement creative; and generating the advertisement creativity of the target product according to the target product information. By applying the scheme provided by the embodiment of the invention, the efficiency of generating the advertisement creativity can be improved.

Description

Advertisement creative generation method and device

Technical Field

The invention belongs to the technical field of information processing, and particularly relates to an advertisement creative generation method and device.

Background

In order to promote own products, an advertiser selects to put advertisements on an advertisement putting platform, before the advertiser puts the advertisements, the advertiser needs to provide product information of the promoted products for the advertisement putting platform, the advertisement putting platform can automatically generate a plurality of advertisement creatives and display the advertisement creatives to the advertiser based on the product information, and the advertiser can select all or part of the advertisement creatives from the advertisement creatives to serve as the actually-put advertisements.

However, when the ad delivery platform generates ad creatives in the above manner, advertisers are required to provide a large amount of product information in order to ensure that a sufficient number of ad creatives are generated. For example, advertisers are required to provide a large amount of product information such as product descriptions, profiles, pictures, videos, etc. And as the number of products to be promoted by the advertiser increases and the number of desired advertisement creatives increases, the amount of product information that the advertiser needs to provide also increases substantially, thus increasing the difficulty of the advertiser in providing the product information, resulting in a longer time for the advertiser to provide the product information, and further resulting in a lower efficiency of generating advertisement creatives.

Disclosure of Invention

The embodiment of the invention aims to provide a method and a device for generating an advertising creative idea so as to improve the efficiency of generating the advertising creative idea. The specific technical scheme is as follows:

in a first aspect, an embodiment of the present invention provides an advertisement creative generation method, where the method includes:

obtaining a target Uniform Resource Locator (URL) of a webpage of a target product;

acquiring the webpage content of the webpage of the target product according to the target URL;

extracting target product information of the target product from the webpage content according to creative items required by the advertising creative;

And generating the advertisement creative of the target product according to the target product information.

In a second aspect, an embodiment of the present invention provides an advertisement creative generating apparatus, including:

the URL obtaining module is used for obtaining a target Uniform Resource Locator (URL) of a webpage of a target product;

the content obtaining module is used for obtaining the webpage content of the webpage of the target product according to the target URL;

the information extraction module is used for extracting target product information of the target product from the webpage content according to creative items required by advertisement creatives;

and the advertisement creative generation module is used for generating the advertisement creative of the target product according to the target product information.

In a third aspect, an embodiment of the present invention provides an electronic device, including a processor, a communication interface, a memory, and a communication bus, where the processor and the communication interface complete communication between the memory and the processor through the communication bus;

a memory for storing a computer program;

a processor, configured to implement the method for generating an advertising creative idea according to the first aspect when executing the program stored in the memory.

In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium, in which a computer program is stored, and the computer program, when executed by a processor, implements the ad creative generation method according to the first aspect.

In a fifth aspect, embodiments of the present invention provide a computer program product containing instructions which, when run on a computer, cause the computer to perform the method for generating an advertising creative according to the first aspect.

As can be seen from the above, when the advertisement creative idea is generated by applying the solution provided by the embodiment of the present invention, the product information of the target product is extracted from the web page content of the web page of the target product according to the creative item required by the advertisement creative idea, so compared with the prior art in which an advertiser provides a large amount of product information when generating the advertisement creative idea, in the solution provided by the embodiment of the present invention, only the advertiser needs to provide the URL (Uniform Resource Locator) of the web page of the target product, and the data volume of the URL is much smaller than that of the product information, so that the time for providing information by the advertiser can be saved, and the efficiency of generating the advertisement creative idea is improved.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the embodiments or the prior art descriptions will be briefly described below.

FIG. 1 is a flowchart illustrating a first method for generating an advertising creative generation according to an embodiment of the present invention;

FIG. 2 is a flowchart illustrating a second method for generating an advertising creative according to an embodiment of the present invention;

FIG. 3 is a flowchart illustrating a third method for generating an advertising creative according to an embodiment of the present invention;

fig. 4 is a schematic structural diagram of a first advertisement creative generation device according to an embodiment of the present invention;

fig. 5 is a schematic structural diagram of a second ad creative generation device according to an embodiment of the present invention;

fig. 6 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.

Detailed Description

The technical solutions in the embodiments of the present invention will be described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

In one embodiment of the present invention, a method for generating an advertising creative includes:

and generating the advertisement creativity of the target product according to the target product information.

As can be seen from the above, when the solution provided by the embodiment of the present invention is applied to generating an advertisement creative, product information of a used target product is extracted from web page content of a web page of the target product according to a creative item required by the advertisement creative, so that compared with the prior art in which an advertiser provides a large amount of product information when generating an advertisement creative, in the solution provided by the embodiment of the present invention, only a URL (Uniform Resource Locator) of the web page of the target product needs to be provided by the advertiser, and the data size of the URL is much smaller than that of the product information, so that the time for the advertiser to provide information can be saved, thereby improving the efficiency of generating the advertisement creative.

The following is a description of related concepts related to embodiments of the present invention.

1. An advertiser: a subject that needs to have an advertisement placed.

2. An advertisement putting platform: and providing a platform of advertisement delivery resources, wherein the platform can select the advertisements to be delivered according to the value representation information of the advertisements required to be delivered by each advertiser and deliver the advertisements.

3. And (3) advertising creativity: and the advertisement putting platform generates each advertisement to be put aiming at the target product promoted by the advertiser.

The following describes the ad creative generation method provided by the embodiment of the present invention in detail through a specific embodiment.

Referring to fig. 1, fig. 1 is a schematic flow chart of a first advertisement creative generation method according to an embodiment of the present invention, where the method includes the following steps S101-S104.

Step S101: and obtaining a target URL of the webpage of the target product.

The target product is a product expected to be promoted by the advertiser on the advertisement putting platform, the webpage of the target product can be a propaganda page used by the advertiser for displaying the target product, a landing page set by the advertiser, and the like, and the landing page is a landing page when the user accesses the portal website of the advertiser and contains detailed information of the target product

Specifically, the target URL may be provided by an advertiser.

Step S102: and obtaining the webpage content of the webpage of the target product according to the target URL.

Specifically, based on the target URL, the web crawler tool may be used to download all the web contents of the web page pointed by the target URL as the web contents of the web page of the target product.

Step S103: and extracting target product information of the target product from the webpage content according to creative items required by the advertising creative.

Since the web page content contains a large amount of information about the target product, the target product information about the target product can be extracted from the web page content according to the creative item required by the advertising creative.

The creative items required for the advertising creative described above are various components of the advertising creative such as title, body, vignette, purchase links, display drawings, display video, etc.

Specifically, the target product information of the target product can be extracted from the web page content according to the creative item required by the advertising creative in the following three embodiments.

In the first embodiment, the target product information matching the creative item may be extracted from a webpage tag in the webpage content.

For example, the content corresponding to the < h1> - < h6> tag in the web page content is often a title, so the content of the tag can be extracted as target product information for the title creative item required by the advertising creative; the content corresponding to the < p > and < span > tags is often textual description and therefore the content of the tags can be extracted as targeted product information for the profile creative item required for the ad creative. In addition, the content corresponding to the < image > tag is often an image, the content corresponding to the < a > tag is often a hyperlink, the content corresponding to the < video > tag is often a video, and the like, and the target product information can be extracted according to the creative item required by the creative advertisement by referring to the above manner, which is not illustrated here.

For convenience of description, the target product information extracted according to a certain creative item will be hereinafter simply referred to as target product information for the creative item.

In the second embodiment, the web page may be subjected to structured parsing, and target product information of a target product is extracted according to a parsing result, which is specifically shown in steps S203-S205 in the embodiment of fig. 2, and will not be described in detail here.

In a third embodiment, auxiliary information for extracting product information may be obtained, and target product information of the target product may be extracted from the web content according to a creative item required for an advertising creative based on the auxiliary information.

Specifically, the auxiliary information may be a keyword of a target product provided by an advertiser, in which case, the target product information of the target product matching the keyword may be extracted from the web page content according to a creative item required by the creative advertisement. For example, if the target product is clothing, the auxiliary information provided by the advertiser may be keywords such as textile, warm-keeping, and pure cotton, so that when extracting the target product information from the web content, information matching the keywords such as textile, warm-keeping, and pure cotton may be extracted as the target product information according to the creative item required by the creative idea of the advertisement.

The auxiliary information may also be information of an information extraction area of a web page of a target product provided by an advertiser, in which case the target product information may be extracted from the content of the web page of the information extraction area according to a creative item required by the creative of the advertisement.

Because the auxiliary information is also referred when the target product information of the target product is extracted from the webpage content according to the creative item required by the advertisement creative, the extracted target product information is matched with the auxiliary information, and the accuracy of the target product information is improved.

Step S104: and generating the advertisement creative of the target product according to the target product information.

Specifically, an advertising creative for a target product may be generated in the following manner.

In one embodiment, each creative item included in the ad creative may not be preset, and based on the target product information for different creative items, the target product information with a correlation degree greater than a first preset threshold is combined according to the correlation degree between the target product information for different creative items to generate an ad creative for the target product. The correlation degree between the target product information may be a similarity degree between feature vectors of the target product information. For example, for target product information in the form of text, feature vectors of the target product information may be extracted, and then similarity between the feature vectors of the target product information may be calculated, and a feature extraction and feature similarity calculation method in the prior art may be specifically used for the feature vector extraction and the similarity between the feature vectors, which is not limited by the present invention.

In another embodiment, each creative item included in the ad creative may not be preset, and based on the target product information for different creative items, the target product information including the same words with a number greater than a preset number is combined to generate the ad creative of the target product. For example, target product information a is obtained for text creative item extraction, target product information B is obtained for brief introduction creative item extraction, and the target product information B and the target product information a contain two same words: and if the preset number is 1, the number of the same words contained in the target product information A and the target product information B is greater than the preset number, and the target product information A and the target product information B can be combined to serve as an advertising creative idea of the target product. Of course, the above examples are merely examples, and more than 2 pieces of targeted product information for different creative items may be combined in the manner described above to generate an advertising creative.

In another embodiment, information matched with each creative item in a preset advertisement template may be selected from the target product information, and the selected information is filled into an area corresponding to each creative item in the advertisement template to generate an advertisement creative of the target product.

The advertisement template is a template which is preset with patterns and is reserved with a region to be filled corresponding to each creative item, and each creative item of different advertisement templates is different. For example, the ad template A reserves areas for a title creative item, a purchase link creative item, and a picture creative item, and the ad template B reserves areas for a title creative item and a profile creative item.

The preset pattern of the advertisement template can be the layout, the size and the like of the area to be filled corresponding to each creative item.

Specifically, the information matched with each creative item in the preset advertisement template may be target product information for which the targeted creative item is consistent with each creative item in the advertisement template.

After the information matched with each creative item in the preset advertisement template is selected, the selected information can be filled into the area corresponding to each creative item in the advertisement template according to a pre-designed pattern, so that the generated advertisement creative is more exquisite and attractive.

As can be seen from the above, when the solution provided by the embodiment of the present invention is applied to generating an advertisement creative, the product information of the used target product is extracted from the web page content of the web page of the target product according to the creative item required by the advertisement creative, so that compared with the prior art in which an advertiser provides a large amount of product information when generating an advertisement creative, the solution provided by the embodiment of the present invention only needs the advertiser to provide the URL of the web page of the target product, and the data size of the URL is much smaller than that of the product information, so that the time for the advertiser to provide information can be saved, thereby improving the efficiency of generating the advertisement creative.

In one embodiment of the invention, the advertiser may provide additional product information to the advertising platform, such as providing additional product descriptions, product pictures, product videos, and the like. In this case, the additional product information provided by the advertiser may also serve as the target product information for the creative advertisement. Thus, the target product information for generating the advertisement creative is richer, and the content richness and the quantity of the advertisement creative generated based on the target product information are improved.

While various ways of extracting the target product information are listed in step S103 in the foregoing embodiment shown in fig. 1, based on a second implementation manner, the embodiment of the present invention provides a specific implementation manner of generating an advertising creative.

Specifically, referring to fig. 2, fig. 2 is a flowchart illustrating a second method for generating an advertising creative generation according to an embodiment of the present invention, where the method includes the following steps S201-S206.

Step S201: and obtaining a target URL of the webpage of the target product.

Step S202: and obtaining the webpage content of the webpage of the target product according to the target URL.

The steps S201 and S202 are the same as the steps S101 and S102 in the embodiment shown in fig. 1, and are not repeated here.

Step S203: and carrying out structured analysis on the webpage content to obtain an analysis result.

Specifically, the web page tags in the web page content can be traversed, text and multimedia data are extracted according to the meanings of different web page tags, the text type and the multimedia data type are further judged according to the type of the web page tags, the extracted text is divided into different types such as a title, a body and a hyperlink, the multimedia data is divided into a picture type and a video type, and thus the obtained analysis result comprises different types of data.

When the type of the text is judged, in addition to judging the type of the text according to the type of the web tag, for example, judging the text in the < p >, < span > tag as the text of the text type, judging the text in the < h1> - < h6> tag as the text of the title type, and the like, the number of characters of the text in the web tag and the web area occupied by the text can be obtained, the text density is calculated according to the number of characters of the text and the web area occupied by the text, and the type of the text is judged by combining the text density, for example, the text density of the text type is often greater than the text density of the text of the title type, so that the text with the text density greater than a second preset threshold value can be judged as the text of the text type. The above is only an example, and other types of texts can also be determined by setting other different thresholds.

Step S204: and aiming at the text creative item required by the advertisement creative, performing natural language processing on the text in the analysis result, and obtaining the target product information of the target product according to the processing result.

The text creative items required by the advertising creative are creative items in the form of text, for example, the creative items such as the title, the introduction, etc. are text creative items. These text creative items may be pre-set.

After the parsing result is obtained, the text corresponding to the text creative item required by the advertisement creative can be selected as the target product information of the target product based on the text type in the parsing result. For example, a text of a title type in the parsing result may be selected as the target product information, and a text of a body type may be selected as the target product information.

In some cases, the text creative item required by the advertisement creative may be further refined, for example, the text creative item may further include more refined creative items such as question and answer, abstract, consultation telephone, store address, and the like, and therefore, in order to extract the target product information for the text creative item from the parsing result, natural language processing needs to be performed on the text in the parsing result to obtain semantic information in the text, and then the target product information for the creative item in the parsing result is determined according to the semantic information.

For example, for a question-answer creative item, semantic analysis needs to be performed on the text in the analysis result, and the text in the question-answer form is determined according to the semantic information obtained through the analysis, and is used as the target product information for the question-answer creative item.

Specifically, the natural language processing may include lexical analysis, syntactic analysis, semantic analysis, and synonym transformation.

In an embodiment, the natural language processing on the text in the parsing result may further include performing sensitive word filtering on the text to filter out preset sensitive words.

Step S205: and performing visual processing on the multimedia data in the analysis result aiming at the multimedia creative item required by the advertisement creative, and obtaining target product information according to the processing result.

The multimedia creative items required for the advertising creative described above can be understood as: creative items in the form of multimedia data, such as the aforementioned presentation diagrams, presentation videos, and the like, are multimedia creative items. The multimedia creative item may be pre-defined.

After the parsing result is obtained, the multimedia data of the multimedia creative item required by the advertising creative can be selected as the target product information of the target product based on the multimedia data type in the parsing result. For example, multimedia data of a picture type in the parsing result may be selected as target product information, multimedia data of a video type may be selected as target product information, and the like.

In some cases, the multimedia creative items required for an advertising creative have formal requirements, e.g., requirements for data size, resolution, etc. In this case, the multimedia data in the parsing result needs to be visually processed, and the target product information of the target product is obtained according to the processing result.

Specifically, the visual processing may include clipping, compressing, watermark removing, and the like, and after the format requirement of the multimedia creative item is met, the processed multimedia data is used as target product information extracted for the multimedia creative item.

In one embodiment, the text in the multimedia data may be recognized by using an OCR (Optical Character Recognition) technique, so as to remove the sensitive words in the multimedia data containing the preset sensitive words.

In an embodiment of the present invention, since some multimedia data may include text, on the basis of this step, the foregoing step S204 may further include the following implementation manners: aiming at the text creative item required by the advertisement creative, the text in the multimedia data in the analysis result can be extracted as target product information by using an OCR technology, then the extracted text is processed by using a natural language, and the creative item aimed at by the target product information is judged according to the processing result.

Step S206: and generating the advertisement creative of the target product according to the target product information.

Step S206 is the same as step S104 in the embodiment shown in fig. 1, and is not described here again.

Therefore, after the webpage content is structurally analyzed, the text and the multimedia data in the analysis result are further processed in a natural language mode and a visual mode, so that based on the processing result, richer and more semantically refined target product information can be extracted according to the text creative item of the advertisement creative, and target product information which meets requirements can be extracted according to the multimedia creative item of the advertisement creative.

After generating an advertising creative in accordance with the embodiment shown in fig. 1, preview information corresponding to the advertising creative may be generated for each advertising creative so as to present the preview information to an advertiser, which will be described in detail below in conjunction with the embodiment shown in fig. 3.

Referring to fig. 3, fig. 3 is a flowchart illustrating a third method for generating an advertising creative according to an embodiment of the present invention, where the method includes the following steps S301-S305.

Step S301: and obtaining a target URL of the webpage of the target product.

Step S302: and obtaining the webpage content of the webpage of the target product according to the target URL.

Step S303: target product information of the target product is extracted from the web page content according to creative items required by the advertising creative.

Step S304: and generating the advertisement creative of the target product according to the target product information.

The steps S301-S304 are the same as the steps S101-S104 in the embodiment shown in FIG. 1, and are not repeated here.

Step S305: and generating preview information corresponding to each advertisement creative, and sending the generated preview information to a client used by the advertiser so that the client displays the received preview information to the advertiser.

Wherein the preview information includes at least one of the following information: the number of intended audience users of an ad creative; expected unit conversion consumption of ad creatives; the number of prospective conversion users of an advertising creative.

The client used by the advertiser is a client used by the advertiser to access the advertisement delivery platform, and can be a software end installed in equipment such as a mobile phone, a desktop computer and a tablet computer.

The preview information is information that can be referred to by an advertiser when selecting an ad creative to be actually delivered.

The concept of each preview information will be explained first.

1. The number of users of the expected audience of the ad creative is as follows: the number of users who are expected to be interested in the ad creative. Often, when a user is interested in an advertisement, a click operation is performed on the advertisement in order to expect to obtain more understanding, and therefore, in one case, the number of users of the expected audience is considered to be the number of users who are expected to create a click operation in the process of actually delivering the advertisement.

2. The expected unit conversion consumption of the ad creative is as follows: unit consumption resulting from performing the set action item for the ad creative is expected in the context of the actual delivery of the ad creative. The operation items set for the advertising creative are preset by the advertiser and represent operations that the advertiser wants the user to see the advertising creative and then execute, and the operation items can be registration, login, order placement and the like. The expected conversion consumption of the ad creative is borne by the advertiser, which can be understood as the consumption that the advertiser needs to pay for each time the set operation items are executed by the user on average.

3. The number of prospective conversion users of the advertising creative is as follows: in the scenario where the ad creative is actually delivered, the number of users who will perform the set operation item on the ad creative is expected.

A specific mode of generating the preview information will be described below.

The number of intended audience users of the ad creative described above for each ad creative may be implemented as follows, steps a-D.

Step A: a first feature vector of the advertising creative is obtained.

Specifically, the manner in which the first feature vector of the advertising creative is obtained can be illustrated by the following three scenarios according to various situations of the creative item included in the advertising creative.

In the first case, the ad creative includes only text creative items. In this case, the first feature vector may be obtained by feature extraction of the text creative item. Specifically, after all texts in the text creative item of the ad creative are obtained, features of the texts may be extracted based on an existing feature vector extraction manner in the prior art, for example, feature vectors in the texts are extracted based on a Doc2Vec algorithm, feature vectors in the texts are extracted based on a Word2Vec algorithm, and the like, and the extracted feature vectors are used as first feature vectors of the ad creative.

In the second case, the ad creative includes only multimedia creative items. In this case, the first feature vector may be obtained by feature extraction of an image of the multimedia creative item. The images in the multimedia creative item may include pictures in the multimedia creative item, or may include screenshots of videos in the multimedia creative item. Specifically, after the images in the multimedia creative item are obtained, the features of the images may be extracted based on a feature vector extraction method existing in the prior art, for example, the feature vectors of the images are extracted based on an edge extraction operator, the feature vectors of the images are extracted based on a texture feature extraction algorithm, the feature vectors of the images are extracted based on a pre-trained convolutional neural network model, and the extracted feature vectors are used as the first feature vectors of the advertising creative.

In a third scenario, the advertising creative includes a text creative item and a multimedia creative item. In this case, after the feature vectors extracted from the text creative item based on the advertising creative and the feature vectors extracted from the multimedia creative item based on the advertising creative are obtained, the two feature vectors may be vector-spliced, and the feature vectors generated after splicing are used as the first feature vectors of the advertising creative.

And B, step B: and obtaining a second feature vector of each existing user of the advertisement putting platform.

Specifically, according to the user portrait of each existing user recorded in advance by the advertisement delivery platform, the interest tag of each existing user is obtained from the user portrait, the feature extraction is further performed on the interest tag, and the extracted feature vector is used as the second feature vector of each existing user. The above method for extracting the feature of the interest tag may refer to the above text feature extraction method, which is not described herein again.

And C: and calculating the expected preference degree of each existing user for the advertising creative based on the first feature vector and the second feature vector of each existing user.

Specifically, the similarity between the first feature vector and the second feature vector of each existing user may be calculated, and the similarity between the second feature vector of each existing user and the first feature vector may be used as the expected preference degree of each existing user for the advertisement creative idea, and the higher the similarity is, the higher the expected preference degree of the user for the advertisement creative idea is. The similarity between the first eigenvector and the second eigenvector may be calculated by using a euclidean distance, a cosine similarity, a pearson correlation coefficient, and the like between the vectors.

Step D: the number of intended audience users for the ad creative is obtained based on the expected likeness of each existing user to the ad creative.

Specifically, the number of users whose expected like degree of the creative is greater than the third preset threshold among the existing users may be counted as the number of users of the expected audience of the creative of the advertisement. The third preset threshold may be set by a worker according to experience.

The number of prospective audience users of the advertising creative can be obtained, the number of the prospective audience users can be understood as the number of the prospective users interested in the advertising creative, therefore, the number of the prospective users interested in the advertising creative can be intuitively known by an advertiser based on the number of the prospective audience users of the advertising creative, and the advertiser is further helped to judge whether to actually deliver the advertising creative or not based on the number of the prospective audience users, the experience of the advertiser when the advertising creative is selected is improved, and the viscosity of the advertiser for delivering the advertisement on an advertising platform is favorably improved.

The expected unit conversion consumption of the ad creative described above for each ad creative may be implemented as follows, steps E-G.

And E, step E: and obtaining the target advertisement similar to the creative of the advertisement from the advertisement delivered by the advertisement delivery platform.

Specifically, the tags of the delivered advertisements can be obtained from the knowledge graph recorded by the advertisement delivery platform, feature extraction is performed on the tags of the delivered advertisements, feature vectors of the delivered advertisements are obtained, the similarity between the feature vectors and the first feature vector of the creative advertisement is calculated, and if the similarity is greater than a fourth preset threshold, the delivered advertisements are target advertisements similar to the creative advertisement. The fourth preset threshold may be set by a worker according to experience. The above feature extraction method and the above similarity calculation method for the tags with advertisements may refer to the above text feature extraction method and the feature similarity calculation method, which are not described herein again.

Step F: and obtaining the unit conversion consumption of the target advertisement.

Wherein the unit conversion consumption is unit consumption generated by an operation item for setting the target advertisement.

Since the target advertisement is delivered on the advertisement delivery platform, the advertisement delivery platform records the unit conversion consumption of the target advertisement, and therefore the unit conversion consumption of the target advertisement recorded by the advertisement delivery platform can be read.

G: based on the unit conversion consumption of the targeted ad, an expected unit conversion consumption of the ad creative is obtained.

Specifically, the expected conversion consumption of units for the ad creative may be obtained in the following manner.

In one embodiment, the average of the unit conversion consumption of the target advertisement may be calculated, and the obtained average may be used as the expected unit conversion consumption of the creative advertisement.

In another embodiment, a weight value may be set for each target advertisement based on the similarity between the target advertisement and the advertisement creative, a weighted average of unit conversion consumption of the target advertisement may be calculated according to the set weight value, and the obtained weighted average may be used as the expected unit conversion consumption of the advertisement creative.

For example, three target ads similar to the ad creative are obtained, namely a target ad a, a target ad B and a target ad C, the unit conversion consumption of the three target ads is 40, 40 and 20 respectively, the similarity between the three target ads and the ad creative is 90%, 60% and 50% respectively, the weight value set for the target ad a is 90%/(90% + 60% + 50%) 9/20, the weight values set for the target ad B and the target ad C are 6/20 and 5/20 respectively, and the expected unit conversion consumption is 40 × 9/20+40 × 6/20+20 × 5/20 — 35.

Therefore, the expected unit conversion consumption of the advertisement creatives can be obtained, so that an advertiser can visually know the expected unit conversion consumption of the advertisement creatives in the actual delivery process, and the advertiser can judge whether to actually deliver the advertisement creatives or not based on the number of the expected unit conversion consumption, the experience of the advertiser when the advertisement creatives are selected is improved, and the viscosity of the advertiser delivering the advertisements on an advertisement delivery platform is favorably improved.

The number of prospective conversion users for each advertising creative described above can be implemented as follows, steps H-J.

Step H: an intended audience user for the ad creative is determined from existing users of the ad delivery platform.

Specifically, the intended audience users may be determined in the following manner.

In one embodiment, a target advertisement similar to the creative idea of the advertisement can be obtained from the advertisement delivered by the advertisement delivery platform, and the specific embodiment is detailed in step E, which is not described herein again; then through reading the information of launching of above-mentioned target advertisement of advertisement putting platform record, confirm above-mentioned target advertisement and produce the user of click operation to this target advertisement user in the process of launching, because above-mentioned user can characterize the audience user of above-mentioned target advertisement, and above-mentioned target advertisement and this advertisement intention are similar, consequently can regard as the anticipated audience user of this advertisement intention with above-mentioned user.

In another embodiment, the existing users with expected like-degree greater than the third preset threshold can be determined as the expected audience users of the ad creative by referring to the method of step a.

Step I: for each expected audience user, the click probability and the conversion probability of the expected audience user on the advertisement creative in the actual delivery scene are predicted.

Wherein, the conversion probability is the probability of the expected audience user generating the preset operation item for the advertisement creative idea.

In this step, a pre-trained double-tower prediction model may be used, and for each expected audience user, the user vector of the expected audience user and the advertisement creative idea vector of the advertisement creative are used as input features of the double-tower prediction model, and the double-tower prediction model may output the click probability and the conversion probability of the expected audience user on the advertisement creative idea in an actual delivery scene.

The user vector of the expected audience user can be obtained by obtaining a user tag from a user portrait of the expected audience user recorded in advance by an advertisement delivery platform, and then performing feature extraction on the user tag to obtain the user vector of the expected audience user.

The following describes how the ad creative vector of the ad creative is obtained.

In one embodiment, the first feature vector of the advertising creative may be used as the advertising creative vector. The manner of obtaining the first feature vector is detailed in step a.

In another embodiment, in addition to the first feature vector of the ad creative idea as the ad creative idea vector, a knowledge map recorded by an ad delivery platform may be used to obtain tags of delivered ads similar to the ad creative idea from the knowledge map, perform feature extraction on the tags of the delivered ads, and use the obtained feature vector of the delivered ads as the ad creative idea vector. The above manner of extracting the feature of each tag that has delivered an advertisement may refer to the above text feature extraction manner, which is not described herein again.

Step J: the number of prospective conversion users for the advertising creative is obtained based on the click probability and conversion probability of each prospective audience user for the advertising creative.

The expected conversion user is an expected audience user which is expected to generate the preset operation item for the target advertisement.

Specifically, for each expected audience user, the product of the click probability and the conversion probability of the expected audience user on the ad creative may be calculated, if the product is greater than a fifth preset threshold, the expected audience user is an expected conversion user, and the number of the expected audience users whose product is greater than the fifth preset threshold is counted as the number of the expected conversion users. The fifth preset threshold may be set by a worker based on experience.

For example, if the click probability and the conversion probability of a desired audience user for the ad creative are 90% and 60%, the product of the click probability and the conversion probability is 90% × 60% > -0.54, and if the fifth preset threshold is 0.5, the product is greater than the fifth preset threshold, the desired audience user is the desired conversion user.

The number of expected conversion users of the advertisement creatives can be obtained, the number of the expected conversion users is indicated under the scene that the advertisement creatives are actually delivered, the number of users expecting to execute the set operation items on the advertisement creatives is expected, therefore, the number of the expected conversion users of the advertisement creatives based on the advertisement creatives can visually know the number of the expected users who can execute the set operation items on the advertisement creatives, and the advertisement creatives can be judged to be actually delivered based on the number of the expected conversion users by the advertisers, the experience of the advertisers when the advertisement creatives are selected is improved, and the viscosity of the advertisements delivered by the advertisers on an advertisement delivery platform is improved.

In one embodiment of the present invention, the preview information may further include expected revenue in case that the operation item set for the ad creative is a purchase or a purchase order.

Specifically, since the number of prospective conversion users of the ad creative idea is determined, in the case that the preset operation item is purchase or order placement, the prospective conversion users can be the prospective purchase or order placement users, and then, the product unit price preset by the advertiser is multiplied by the prospective purchase or order placement users, and the obtained product is the prospective profit of the advertiser. Therefore, the advertiser can know whether the advertisement creative is expected by the advertiser for the advertiser to achieve the self income or not under the actual delivery scene of the advertisement creative, and the advertiser is helped to judge whether the advertisement creative is actually delivered or not based on the expected income, so that the experience of the advertiser is improved, and the viscosity of delivering the advertisement on an advertisement delivery platform by the advertiser is favorably improved.

By above, after each advertisement intention is generated, in order to facilitate a user to select an actually delivered advertisement intention from the advertisement originality, corresponding preview information can be generated for each advertisement intention, so that an advertiser can visually acquire each item of preview information of each advertisement intention, and then can decide which advertisement originality to actually deliver based on the preview information, thereby being beneficial to improving the credibility of the advertiser to an advertisement delivery platform and the experience of the advertiser when selecting the advertisement intention, and being beneficial to improving the viscosity of the advertiser delivering the advertisement on the advertisement delivery platform.

In one embodiment of the invention, ad creatives for which the number of prospective conversion users is less than the preset number of users may also be deleted.

Specifically, the number of the above-described prospective conversion users can be determined in the following manner.

In one embodiment, the preset number of users may be set by the advertiser in combination with the promoted target product and the operation items set for the target product, and based on the expectation of the advertiser. For example, in the case where the product promoted by the advertiser is a luxury and the operation item set is a purchase, the number of the above-mentioned prospective conversion users set by the advertiser may be 10, 20, or the like; in the case where the product promoted by the advertiser is food and the set operation item is purchase, the number of the prospective conversion users set by the advertiser may be 100, 200, or the like.

In another embodiment, the preset number of users may also be a fixed value set by an operator according to experience. For example, the above-mentioned preset number of users is set to 1, and so on.

Therefore, the method can help the advertiser to filter out the advertisement creativity which does not meet the preset requirement, is beneficial to reducing the time for selecting the advertisement from the advertiser and the advertisement creativity, improves the experience of the advertiser, and is beneficial to improving the viscosity of the advertiser for putting the advertisement on the advertisement putting platform.

Corresponding to the method for generating the advertisement creative idea, the embodiment of the invention also provides a device for generating the advertisement creative idea.

Referring to fig. 4, fig. 4 is a schematic structural diagram of a first advertisement creative generation apparatus according to an embodiment of the present invention, where the apparatus includes the following

modules

401 and 404.

A URL obtaining module 401, configured to obtain a target uniform resource locator URL of a webpage of a target product;

a content obtaining module 402, configured to obtain, according to the target URL, a web page content of a web page of the target product;

an information extraction module 403, configured to extract target product information of the target product from the web page content according to a creative item required by an advertisement creative;

and an advertising creative generation module 404, configured to generate an advertising creative of the target product according to the target product information.

In an embodiment of the present invention, the information extracting module 403 is specifically configured to obtain auxiliary information for extracting product information; and extracting target product information of the target product from the webpage content according to creative items required by the advertising creative based on the auxiliary information.

In an embodiment of the present invention, the information extraction module 403 is specifically configured to perform structured parsing on the web page content to obtain a parsing result; performing natural language processing on the text in the analysis result aiming at the text creative item required by the advertisement creative, and obtaining the target product information of the target product according to the processing result; and performing visual processing on the multimedia data in the analysis result aiming at the multimedia creative item required by the advertisement creative, and obtaining the target product information according to the processing result.

In an embodiment of the present invention, the advertisement creative generation module 404 is specifically configured to select information matching with creative items in a preset advertisement template from the target product information; and filling the selected information into the areas corresponding to the creative items in the advertisement template to generate the advertisement creative of the target product.

After the information matched with each creative item in the preset advertisement template is selected, the selected information can be filled into the area corresponding to each creative item in the advertisement template according to a pre-designed style, so that the generated advertisement creative is more exquisite and attractive.

Referring to fig. 5, fig. 5 is a schematic structural diagram of a second advertisement creative generation device according to an embodiment of the present invention, where the device includes the following

modules

501 and 505.

A URL obtaining module 501, configured to obtain a target uniform resource locator URL of a webpage of a target product;

a content obtaining module 502, configured to obtain, according to the target URL, a web page content of a web page of the target product;

an information extraction module 503, configured to extract target product information of the target product from the web page content according to a creative item required by an advertisement creative;

An advertising creative generation module 504, configured to generate an advertising creative for the target product according to the target product information.

A preview information generating module 505, configured to generate preview information corresponding to each ad creative after the ad creative generating module 504 generates an ad creative, and send the generated preview information to a client used by an advertiser, so that the client displays the received preview information to the advertiser;

wherein the preview information includes at least one of:

the number of intended audience users of an ad creative;

expected unit conversion consumption of ad creatives;

the number of prospective conversion users of the ad creative.

Therefore, after each advertisement creative is generated, in order to facilitate a user to select an actually delivered advertisement creative from the advertisement creatives, corresponding preview information can be generated for each advertisement creative, so that an advertiser can visually acquire each preview information of each advertisement creative, and then can determine which advertisement creative is actually delivered based on the preview information, thereby being beneficial to improving the trust degree of the advertiser on an advertisement delivery platform and the experience of the advertiser when selecting the advertisement creative, and being beneficial to improving the viscosity of the advertiser delivering the advertisement on the advertisement delivery platform.

In one embodiment of the invention, for each ad creative, the number of intended audience users for the ad creative is generated as follows:

obtaining a first feature vector of the advertising creative; obtaining a second feature vector of each existing user of the advertisement putting platform; calculating the expected preference degree of each existing user for the advertisement creative based on the first feature vector and the second feature vector of each existing user; the number of intended audience users for the ad creative is obtained based on the expected likeness of each existing user to the ad creative.

The number of prospective audience users of the advertising creative can be obtained, the number of the prospective audience users can be understood as the number of the prospective users interested in the advertising creative, therefore, the number of the prospective users interested in the advertising creative can be intuitively known by an advertiser based on the number of the prospective audience users of the advertising creative, and therefore the advertiser can judge whether to actually deliver the advertising creative or not based on the number of the prospective audience users, the experience of the advertiser when the advertising creative is selected is improved, and the method is favorable for improving the viscosity of the advertiser to deliver the advertisement on an advertising platform.

In one embodiment of the invention, for each ad creative, the expected unit conversion consumption for that ad creative is generated as follows:

obtaining a target advertisement similar to the creative of the advertisement from the advertisements released by the advertisement releasing platform; obtaining unit conversion consumption of the target advertisement, wherein the unit conversion consumption is unit consumption generated by executing set operation items on the target advertisement; based on the unit conversion consumption of the targeted advertisement, an expected unit conversion consumption of the ad creative is obtained.

Therefore, the expected unit conversion consumption of the advertisement creative idea can be obtained, so that an advertiser can visually know the expected unit conversion consumption of the advertisement creative idea in the actual delivery process, and the advertiser can judge whether to actually deliver the advertisement creative idea based on the number of the expected unit conversion consumption, the experience of the advertiser when the advertisement creative is selected is improved, and the improvement of the viscosity of the advertiser for delivering the advertisement on an advertisement delivery platform is facilitated.

In one embodiment of the invention, for each ad creative, the number of prospective conversion users for the ad creative is generated as follows:

Determining expected audience users of the advertisement creative idea from existing users of the advertisement delivery platform; predicting the click probability and the conversion probability of the expected audience user to the advertisement creative idea under the actual delivery scene aiming at each expected audience user, wherein the conversion probability is the probability of the expected audience user to generate a preset operation item on the advertisement creative idea; and obtaining the number of expected conversion users of the advertising creative based on the click probability and the conversion probability of each expected audience user to the advertising creative, wherein the expected conversion users are the expected audience users expected to generate the preset operation items for the target advertisement.

Therefore, the number of the expected conversion users of the advertisement creative idea can be obtained, the number of the expected conversion users represents the number of users expecting to execute the set operation items on the advertisement creative idea under the actual delivery scene of the advertisement creative idea, so that the number of the expected conversion users of the advertiser based on the advertisement creative idea can intuitively know the number of the expected users expecting to execute the set operation items on the advertisement creative idea, the advertiser is further helped to judge whether to actually deliver the advertisement creative idea or not based on the number of the expected conversion users, the experience of the advertiser when the advertisement creative idea is selected is improved, and the viscosity of delivering the advertisement on an advertisement delivery platform is favorably improved.

In one embodiment of the invention, the apparatus further comprises:

and the creative deletion module is used for deleting the advertising creatives of which the number of the expected conversion users is less than the number of the preset users.

Therefore, the method can help the advertiser to filter out the advertisement originality which is not in line with the requirement of the advertiser, is beneficial to reducing the time for selecting the advertisement originality of the advertiser, improves the experience of the advertiser, and is beneficial to improving the viscosity of the advertiser for putting the advertisement on the advertisement putting platform.

An embodiment of the present invention further provides an electronic device, as shown in fig. 6, including a processor 601, a communication interface 602, a memory 603, and a communication bus 604, where the processor 601, the communication interface 602, and the memory 603 complete mutual communication through the communication bus 604,

a memory 603 for storing a computer program;

the processor 601 is configured to implement the steps of the ad creative generation method provided by the above-mentioned method embodiments when executing the program stored in the memory 603.

The communication bus mentioned in the above terminal may be a Peripheral Component Interconnect (PCI) bus, an Extended Industry Standard Architecture (EISA) bus, or the like. The communication bus may be divided into an address bus, a data bus, a control bus, etc. For ease of illustration, only one thick line is shown, but this does not mean that there is only one bus or one type of bus.

The communication interface is used for communication between the terminal and other devices.

The Memory may include a Random Access Memory (RAM), and may also include a non-volatile Memory (non-volatile Memory), such as at least one disk Memory. Alternatively, the memory may be at least one memory device located remotely from the processor.

The Processor may be a general-purpose Processor, including a Central Processing Unit (CPU), a Network Processor (NP), and the like; the Integrated Circuit may also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field Programmable Gate Array (FPGA) or other Programmable logic device, discrete Gate or transistor logic device, or discrete hardware components.

In yet another embodiment of the present invention, a computer-readable storage medium is further provided, in which a computer program is stored, and the computer program, when executed by a processor, implements the advertising creative generation method in any of the above embodiments.

In yet another embodiment provided by the present invention, there is also provided a computer program product containing instructions which, when run on a computer, cause the computer to perform the advertising creative generation method of any of the above embodiments.

In the above embodiments, the implementation may be wholly or partially realized by software, hardware, firmware, or any combination thereof. When implemented in software, may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When loaded and executed on a computer, cause the processes or functions described in accordance with the embodiments of the invention to occur, in whole or in part. The computer may be a general purpose computer, a special purpose computer, a network of computers, or other programmable device. The computer instructions may be stored in a computer readable storage medium or transmitted from one computer readable storage medium to another computer readable storage medium, for example, the computer instructions may be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, Digital Subscriber Line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device, such as a server, a data center, etc., that incorporates one or more of the available media. The usable medium may be a magnetic medium (e.g., floppy Disk, hard Disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., Solid State Disk (SSD)), among others.

It should be noted that, in this document, relational terms such as first and second, and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element.

All the embodiments in the present specification are described in a related manner, and the same and similar parts among the embodiments may be referred to each other, and each embodiment focuses on the differences from the other embodiments. In particular, for the embodiments of the apparatus, the electronic device, and the computer-readable storage medium, since they are substantially similar to the embodiments of the method, the description is simple, and for the relevant points, reference may be made to the partial description of the embodiments of the method.

The above description is only for the preferred embodiment of the present invention, and is not intended to limit the scope of the present invention. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention shall fall within the protection scope of the present invention.

Claims

1. An advertising creative generation method, the method comprising:

2. The method of claim 1, wherein extracting target product information for the target product from the web content as required by the creative item of the advertisement, comprises:

obtaining auxiliary information for extracting product information;

and extracting target product information of the target product from the webpage content according to creative items required by the advertising creative based on the auxiliary information.

3. The method of claim 1, wherein extracting target product information for the target product from the web content as required by the creative item of the advertisement, comprises:

Carrying out structured analysis on the webpage content to obtain an analysis result;

aiming at a text creative item required by the advertisement creative, performing natural language processing on the text in the analysis result, and obtaining target product information of the target product according to the processing result;

and performing visual processing on the multimedia data in the analysis result aiming at the multimedia creative item required by the advertisement creative, and obtaining the target product information according to the processing result.

4. The method of claim 1, wherein generating an advertising creative for the target product based on the target product information comprises:

selecting information matched with each creative item in a preset advertisement template from the target product information;

and filling the selected information into the area corresponding to each creative item in the advertisement template to generate the advertisement creative of the target product.

5. The method of any of claims 1-4, after generating the advertising creative, further comprising:

generating preview information corresponding to each advertisement creative, and sending the generated preview information to a client used by an advertiser so that the client displays the received preview information to the advertiser;

Wherein the preview information includes at least one of:

the number of intended audience users of an ad creative;

expected unit conversion consumption of ad creatives;

the number of prospective conversion users of the ad creative.

6. The method of claim 5,

for each ad creative, the number of intended audience users for the ad creative is generated as follows:

obtaining a first feature vector of the advertising creative;

obtaining a second feature vector of each existing user of the advertisement putting platform;

calculating the expected preference degree of each existing user to the advertisement creative idea based on the first feature vector and the second feature vector of each existing user;

the number of intended audience users for the advertising creative is obtained based on the expected likeness of each existing user to the advertising creative.

7. The method of claim 5,

for each ad creative, the expected unit conversion consumption for the ad creative is generated as follows:

obtaining a target advertisement similar to the creative of the advertisement from the advertisements released by the advertisement releasing platform;

obtaining unit conversion consumption of the target advertisement, wherein the unit conversion consumption is unit consumption generated by executing set operation items on the target advertisement;

Based on the unit conversion consumption of the target advertisement, an expected unit conversion consumption of the ad creative is obtained.

8. The method of claim 5,

for each ad creative, the number of prospective conversion users for the ad creative is generated as follows:

determining expected audience users of the advertisement creative from existing users of the advertisement delivery platform;

predicting the click probability and the conversion probability of the expected audience user on the advertisement creative in the actual delivery scene aiming at each expected audience user, wherein the conversion probability is the probability of the expected audience user on generating a preset operation item on the advertisement creative;

and obtaining the number of expected conversion users of the advertisement creative based on the click probability and the conversion probability of each expected audience user on the advertisement creative, wherein the expected conversion users are the expected audience users expected to generate the preset operation items on the target advertisement.

9. The method of claim 5, further comprising:

and deleting the advertisement creatives of which the number of the expected conversion users is less than the preset number of users.

10. An advertising creative generation apparatus, the apparatus comprising:

The URI obtaining module is used for obtaining a target Uniform Resource Locator (URL) of a webpage of a target product;

11. An electronic device is characterized by comprising a processor, a communication interface, a memory and a communication bus, wherein the processor and the communication interface are used for realizing mutual communication by the memory through the communication bus;

a memory for storing a computer program;

a processor for implementing the method steps of any of claims 1-9 when executing a program stored in the memory.

12. A computer-readable storage medium, characterized in that a computer program is stored in the computer-readable storage medium, which computer program, when being executed by a processor, carries out the method steps of any one of the claims 1-9.