CN108681920A - The method and system launched based on advertising creative optimization advertisement - Google Patents

Abstract

Embodiments of the present invention provide a method and system for optimizing advertisement placement based on advertisement creativity, and relate to a method for optimizing advertisement placement in an Internet environment. The method includes: preset the corresponding relationship between the marketing target of the advertising delivery platform and the user behavior on the advertising delivery platform; according to the preset corresponding relationship and different marketing objectives, use the historical user behavior on the advertising delivery platform to train the estimation model, and obtain the corresponding relationship The target conversion rate estimation model corresponding to the marketing target; according to the target conversion rate estimation model corresponding to the marketing target, the bidding coefficient of the advertisement is calculated; according to the bidding coefficient of each advertisement and the preset bidding rules, each The final bid for an ad, which will be shown for the winning ad. It realizes the personalized optimization of advertising delivery according to the marketing goals selected by advertisers, realizes the balance between the income of the advertising delivery platform and the interests of advertisers, optimizes advertising delivery, and saves costs on the premise of ensuring the effectiveness of advertising delivery.

Description

Method and system for optimizing advertisement placement based on advertisement creativity

technical field

The invention relates to an Internet advertisement delivery method, in particular to a method for optimizing advertisement delivery based on the advertisement creative itself and feedback from different user behaviors.

Background technique

Internet advertising delivery is an advertising delivery method that puts the advertising materials that advertisers want to deliver to the client through the advertising delivery system, and shows the user to read. It benefits from the development of Internet big data, user portraits, and efficient computing capabilities. The performance of the ad is improved after it is launched. With the progress of the times, and the enrichment, diversification, and refinement of advertising demands, advertisers have increasingly strong demand for advertising effect data. There is a significant relationship between different user behaviors in different scenarios and advertising demands. Therefore, according to user Advertisements can be delivered based on the probability of occurrence of different behaviors of users, which can further improve the effectiveness of advertisements. At the same time, the feedback of these user behaviors can be used by advertisers to further dig out rich content and provide deeper commercial value.

In the prior art, there are mainly two methods of delivering advertisements to specific users based on different user behavior feedbacks: the first kind of scheme is the experience delivery method, that is, the user behavior feedback is analyzed before the advertisement is placed to determine The delivered materials meet the expected user feedback results and maximize the delivery revenue; the second type is the post-experience delivery billing method, that is, the advertiser specifies the target users for delivery, and the delivery platform charges the conversion volume to optimize the delivery. Advertisement delivery effect; the third category is target optimization delivery, that is, the platform selects users with the strongest target feedback when delivering advertisements. By assigning different weights to different user behaviors or setting different thresholds for each user behavior, the purpose of effect optimization is achieved.

In the process of realizing the present invention, the inventor found that there are at least the following problems in the prior art: the first type of method relies heavily on advertising creativity, and it is difficult to guarantee the effect of advertising; the second type of method ignores the interactive behavior of non-target users It is too rough to develop new users; the third method: divide different weights for different user behaviors and select users with the strongest target feedback. This method is not enough to describe the feedback of different users on specific advertising ideas. Different thresholds are set for different user behaviors, which affect the progress of advertising delivery, without taking into account the different effects of different creative ideas of the advertising itself.

Contents of the invention

The embodiment of the present invention provides a method and system for users to optimize advertisement placement based on the context of advertisement creativity itself according to marketing goals, which can optimize the effect of advertisement placement according to the creative situation of the advertisement itself and according to the marketing goal selected by the advertiser, while ensuring the placement of advertisements schedule.

On the one hand, an embodiment of the present invention provides a method for optimizing advertisement placement based on advertisement creativity, the method comprising:

Preset the corresponding relationship between the marketing objectives provided by the advertising delivery platform and the user behavior on the advertising delivery platform;

According to the preset corresponding relationship and different marketing goals, the user's historical behavior on the advertising platform is used to perform prediction model training to obtain a target conversion rate prediction model corresponding to the marketing goal. User historical behavior includes: the behavior of all users on the advertising delivery platform mentioned in the first fixed time period set before the current time;

Calculate the bidding coefficient of the advertisement according to the marketing objective selected when the advertisement is launched, and the target conversion rate estimation model corresponding to the marketing objective;

According to the bidding coefficient of each advertisement and the preset bidding rules, the final bid of each advertisement is obtained, and the winning advertisement is exposed to users who meet the targeting conditions selected when the advertisement is placed.

On the other hand, an embodiment of the present invention provides a system for optimizing advertisement placement based on advertisement creativity, and the system includes:

The preset unit is used to preset the corresponding relationship between the marketing objectives provided by the advertising delivery platform and the user behavior on the advertising delivery platform;

The training unit is used to perform prediction model training based on the preset corresponding relationship and different marketing goals based on user historical behavior on the advertising delivery platform to obtain a target conversion rate prediction model corresponding to the marketing goal. The historical user behavior on the advertising delivery platform includes: the behavior of all users on the advertising delivery platform in the first fixed time period set before the current time;

A calculation unit, configured to calculate the bidding coefficient of the advertisement according to the marketing target selected when the advertisement is launched, and the target conversion rate estimation model corresponding to the marketing target;

The bidding unit is used to obtain the final bid of each advertisement according to the bidding coefficient of each advertisement and the preset bidding rules, and expose the winning advertisement to users who meet the targeting conditions selected when the advertisement is placed.

The above technical solution has the following beneficial effects: because the corresponding relationship between the marketing target provided by the advertising delivery platform and the user behavior on the advertising delivery platform is adopted, and according to the preset corresponding relationship, the historical behavior of the user on the advertising delivery platform is used to train The technical means of predicting the model, so the technical effect of obtaining the target conversion rate forecasting model corresponding to the marketing target is achieved; because the marketing target selected according to the advertising time and the target forecasting model corresponding to the marketing target are adopted, Calculate the bidding coefficient of the advertisement, and calculate the final bid of the advertisement according to the bidding coefficient of the advertisement and the preset bidding rules, so as to achieve the marketing goal selected when placing the advertisement, as well as the contextual user of each advertisement creative Feedback behavior to optimize the technical effect of advertising, while ensuring the progress of advertising.

Description of drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings that need to be used in the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are only These are some embodiments of the present invention. Those skilled in the art can also obtain other drawings based on these drawings without creative work.

Fig. 1 is a flow chart of a method for optimizing advertisement placement based on advertisement creativity in Embodiment 1 of the present invention;

Fig. 2 is a schematic structural diagram of a system for optimizing advertisement placement based on advertisement creativity in Embodiment 1 of the present invention;

3 is a flowchart of a method for optimizing advertisement delivery based on advertisement creativity in Embodiment 2 of the present invention;

4 is a schematic structural diagram of a system for optimizing advertisement placement based on advertisement creativity in Embodiment 2 of the present invention;

Fig. 5 is a sub-flow chart of obtaining the target conversion rate estimation model corresponding to the marketing target according to the embodiment of the present invention;

Fig. 6 is a schematic structural diagram of a training unit according to an embodiment of the present invention;

7 is a schematic diagram of a target conversion rate estimation model corresponding to a marketing target obtained in an embodiment of the present invention;

Fig. 8 is a schematic diagram of obtaining a target conversion reference value according to an embodiment of the present invention.

Detailed ways

The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some, not all, embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without making creative efforts belong to the protection scope of the present invention.

Please refer to FIG. 1. FIG. 1 is a flowchart of a method for optimizing advertisement placement based on advertisement creativity in Embodiment 1 of the present invention;

101. Preset the corresponding relationship between the marketing objectives provided by the advertising delivery platform and the user behavior on the advertising delivery platform;

102. According to the preset corresponding relationship and different marketing objectives, the user's historical behavior on the advertising delivery platform is used to perform estimation model training to obtain a target conversion rate estimation model corresponding to the marketing objective, and the advertising delivery The historical behavior of users on the platform includes: the behavior of all users on the advertising delivery platform in the first fixed time period set before the current time;

103. Calculate the bidding coefficient of the advertisement according to the marketing goal selected when the advertisement is launched, and the target conversion rate estimation model corresponding to the marketing goal;

104. According to the bidding coefficient of each advertisement and the preset bidding rules, the final bid of each advertisement is obtained, and the winning advertisement is exposed to the users who meet the targeting conditions selected when placing the advertisement.

Please refer to FIG. 3. FIG. 3 is a flowchart of a method for optimizing advertisement placement based on advertisement creativity in Embodiment 2 of the present invention;

101. Preset the corresponding relationship between the marketing objectives provided by the advertising delivery platform and the user behavior on the advertising delivery platform;

102. According to the preset corresponding relationship and different marketing objectives, the user's historical behavior on the advertising delivery platform is used to perform estimation model training to obtain a target conversion rate estimation model corresponding to the marketing objective, and the advertising delivery The historical behavior of users on the platform includes: the behavior of all users on the advertising delivery platform mentioned in the first fixed time period set before the current time;

105. For the current user of the advertisement delivery platform, match according to the targeting condition selected when the advertisement is placed, and filter out the complete set of advertisements to be exposed, and the targeting condition selected during the advertisement delivery is the attribute information of the pre-advertising user;

106. Calculate the bidding coefficient of each advertisement according to the marketing target selected when each advertisement in the complete set of advertisements to be exposed is launched, and the target conversion rate estimation model corresponding to the marketing target;

107. Obtain the final bid of each advertisement in the full set of advertisements to be exposed according to the bidding coefficient of each advertisement and the preset bidding rules, and deliver the winning advertisement to the current user.

Preferably, the marketing objectives provided by the preset advertising delivery platform include: increasing the number of fans, increasing customer inquiries, and increasing brand awareness;

Preferably, user behaviors on the advertising delivery platform include: following, reposting, commenting, liking, bookmarking, clicking on the avatar, and clicking on the details page;

Preferably, the correspondence between the marketing objectives provided by the preset advertising delivery platform and user behavior on the advertising delivery platform includes: increasing the number of fans corresponding to attention behavior, increasing customer consultation and comments, and corresponding to clicking on the detailed page behavior, Increasing brand awareness corresponds to following, forwarding, commenting, liking, saving, clicking on the profile picture, and clicking on the details page.

Please refer to FIG. 5. FIG. 5 is a sub-flow chart of obtaining the target conversion rate prediction model corresponding to the marketing target according to the embodiment of the present invention;

102.1. Collect the behaviors of all users on the advertising delivery platform within the set first fixed time period, and use the presence or absence of user behaviors corresponding to different preset marketing objectives as training samples;

102.2. Train the known click through rate prediction model offline according to the training samples, and obtain the target conversion rate prediction model corresponding to the marketing target.

Preferably, calculating the bidding coefficient of the advertisement specifically includes:

According to the preset marketing target when the advertisement is launched, the target conversion rate on the current traffic is estimated by using the target conversion rate estimation model corresponding to the marketing target;

Store the target conversion rate on the current traffic in the context of the corresponding advertisement, calculate the reference value of the target conversion rate in the second fixed time period set before the current time, and the value of the second fixed time period A value less than the first fixed time period;

According to the marketing target optimization function preset by the advertising delivery platform, as well as the target conversion rate and the reference value of the target conversion rate, calculate the bidding coefficient of each advertisement:

Preferably, the winning advertisement refers to the advertisement with the highest final bid;

Preferably, the calculation method of the final bid of the advertisement specifically includes:

According to the value function, the value of the bidding coefficient of the advertisement is controlled within a predetermined value range, and the revised bidding coefficient is obtained;

The product of the preset delivery price of the ad and the revised bidding coefficient is used as the final bid of the ad.

Please refer to FIG. 2. FIG. 2 is a schematic structural diagram of a system for optimizing advertisement placement based on advertisement creativity in Embodiment 1 of the present invention;

The preset unit 21 is used to preset the corresponding relationship between the marketing target provided by the advertisement delivery platform and the user behavior on the advertisement delivery platform;

The training unit 22 is used to perform prediction model training based on the preset corresponding relationship and different marketing goals based on user historical behavior on the advertising delivery platform to obtain a target conversion rate prediction model corresponding to the marketing goal, The historical behavior of users on the advertising delivery platform includes: the behavior of all users on the advertising delivery platform in the first fixed time period set before the current time;

The first calculation unit 23 is configured to calculate the bidding coefficient of the advertisement according to the marketing target selected when the advertisement is launched, and the target conversion rate estimation model corresponding to the marketing target;

The first bidding unit 24 is used to obtain the final bid of each advertisement according to the bidding coefficient of each advertisement and the preset bidding rules, and expose the winning advertisement to the users who meet the targeting conditions selected when placing the advertisement.

Please refer to FIG. 4. FIG. 4 is a schematic structural diagram of a system for optimizing advertisement placement based on advertisement creativity in Embodiment 2 of the present invention;

The preset unit 21 is used to preset the corresponding relationship between the marketing target provided by the advertisement delivery platform and the user behavior on the advertisement delivery platform;

The training unit 22 is used to perform prediction model training based on the preset corresponding relationship and different marketing goals based on user historical behavior on the advertising delivery platform to obtain a target conversion rate prediction model corresponding to the marketing goal, The historical behavior of users on the advertising delivery platform includes: the behavior of all users on the advertising delivery platform in the first fixed time period set before the current time;

The screening unit 25 is used to match the current user of the advertisement delivery platform according to the targeting condition selected when the advertisement is delivered, and filter out the complete collection of advertisements to be exposed. The targeting condition selected when the advertisement is delivered is the attribute information of the pre-advertising user ;

The first calculation unit 23 is also used to calculate the bidding coefficient of each advertisement according to the marketing target selected when each advertisement in the full set of advertisements to be exposed is placed, and the target conversion rate estimation model corresponding to the marketing target;

The first bidding unit 24 is also used to obtain the final bid of each advertisement in the full set of advertisements to be exposed according to the advertisement bidding coefficient and the preset bidding rules, and deliver the winning advertisement to the current user.

Preferably, in the preset unit 21, the marketing goals provided by the preset advertising delivery platform include: increasing the number of fans, increasing customer inquiries, and increasing brand awareness;

Preferably, in the preset unit 21, user behaviors on the advertising delivery platform include: following, reposting, commenting, liking, bookmarking, clicking on the avatar, and clicking on the details page;

Preferably, in the preset unit 21, the preset relationship between the marketing objectives provided by the advertising delivery platform and user behavior on the advertising delivery platform includes: increasing the number of fans corresponding to user attention behaviors, increasing customer consultation and user behavior Comments correspond to the behavior of clicking on the details page, and increasing brand awareness corresponds to user attention, reposting, commenting, likes, favorites, clicking on avatars, and clicking on the details page.

Please refer to FIG. 6, which is a schematic structural diagram of a training unit according to an embodiment of the present invention;

The collection module 221 is used to collect all user behaviors on the advertising delivery platform within the set first fixed time period, and obtain the presence or absence of user behaviors corresponding to different preset marketing objectives as training samples;

The training module 222 is configured to obtain a target conversion rate prediction model corresponding to the marketing target based on a click-through rate prediction model known off-line for training samples.

Preferably, in the first calculation unit 23, the module for calculating the advertising bidding coefficient specifically includes:

The first calculation sub-module is used for estimating the target conversion rate on the current traffic by using the target conversion rate estimation model corresponding to the marketing target according to the preset marketing target when the advertisement is placed;

The second calculation sub-module is used to store the target conversion rate on the current traffic in the context of the corresponding advertisement, and calculate the reference value of the target conversion rate in the second fixed time period set before the current time, said The value of the second fixed time period is less than the value of the first fixed time period;

The third calculation sub-module is used to calculate the bidding coefficient of each advertisement according to the marketing target optimization function preset by the advertising delivery platform, as well as the target conversion rate and the reference value of the target conversion rate:

Preferably, the winning advertisement refers to the advertisement with the highest final bid;

Preferably, in the first bidding unit 24, the calculation module for the final bid of the advertisement specifically includes:

The value acquisition sub-module is used to control the numerical value of the bid coefficient of the advertisement within a predetermined numerical range according to the value function, so as to obtain the revised bid coefficient;

The bidding sub-module is used to use the product of the preset delivery price of the advertisement and the revised bidding coefficient as the final bidding price of the advertisement.

The above technical solution has the following technical effects, because the presence or absence of user behavior corresponding to the marketing target is used as an observation sample, and the user behavior of the advertising delivery platform is trained for a period of time to obtain the target conversion rate prediction model corresponding to the marketing target , so different marketing goals are achieved, different user behaviors are used to train the model separately, and the technical effect of higher accuracy is achieved; because the target conversion rate under the current traffic is stored in the context of the advertisement, the reference of the target conversion rate is obtained Value, the technical means of obtaining the bidding coefficient of the advertisement, so the technical effect of optimizing the advertisement placement according to the marketing target selected when placing the advertisement and the feedback from the user in the creative context of the advertisement itself is achieved; because the second fixed value set by the calculation is adopted The technical means of the reference value of the target conversion rate within the time period, the set value of the second fixed time period is smaller than the set value of the first fixed time period, so that different advertisements do not affect each other, the optimal method is equivalent to Personalized scoring is carried out for each ad, and the reference value of the target conversion rate uses the feedback of the ad creative itself, which not only optimizes, but also does not affect the progress of ad delivery. Because of the adoption of the technical means of limiting the value of the bidding coefficient to a predetermined value range, it achieves a more balanced interest consideration in terms of system revenue and advertiser appeals, and the adjustment of the revised bidding coefficient is more intuitive and can achieve system revenue A win-win situation for advertisers and advertisers to reduce costs.

The technical solution is further elaborated below in conjunction with specific application examples:

According to the self-characteristics of the advertising delivery platform and the different marketing needs of different advertisers, the advertising delivery system provides corresponding optional marketing objectives, such as increasing the number of fans, increasing customer inquiries, increasing brand awareness, etc. When advertisers place advertisements, they can choose the marketing goals provided by the advertising system according to their own demands, and the advertising background will optimize the delivery of corresponding marketing goals based on the marketing goals selected by the advertisers and combined with their own characteristics. How to optimize the entire process of advertising delivery based on the marketing objectives selected by all advertisements placed on the advertising platform is called multi-target optimization. The detailed process is as follows:

The corresponding relationship between marketing goals and user behaviors is preset on the advertising delivery platform. A many-to-many relationship is established by combining user behavior on the advertising platform with the marketing objectives provided by the platform. The marketing goals provided by the preset advertising platform include: increase the number of fans, increase customer consultation, and increase brand awareness; user behavior on the advertising platform, including: follow, forward, comment, like, favorite, click on the profile picture, click on the detail page ; Preset the corresponding relationship between the marketing objectives provided by the advertising platform and user behavior on the advertising platform, including: increasing the number of fans corresponding to user attention behavior, increasing customer consultation and user comments, and corresponding to the behavior of clicking on the details page, increasing the brand Popularity corresponds to the behaviors of users following, reposting, commenting, liking, saving, clicking on the profile picture, and clicking on the details page. When advertisers place advertisements, they choose the marketing goals they focus on, and choosing the marketing goals they focus on is single. In other words, advertisers can only choose the core marketing goals they care about the most. For example, the core appeal of Internet celebrities’ advertising is to increase fans quantity.

Please refer to FIG. 7. FIG. 7 is a schematic diagram of the target conversion rate estimation model corresponding to the marketing target according to the embodiment of the present invention; the historical data of all users on the advertising platform are collected, and the historical data of the users is before the current time. All behaviors of all users on the advertising delivery platform in the first fixed period of time are usually collected, and all behaviors of users on the advertising delivery platform in the past 30 days are usually collected. For different marketing objectives, use whether there is any interaction on this objective as a training sample, as shown in the data in the small box on the left side of Figure 7. This data is the presence or absence of user behavior interaction corresponding to the marketing objective. Different marketing The corresponding value for the target will be different. For different marketing objectives, use whether there is any action on the objective as a training sample, and train the click-through rate prediction model offline. Obtain target conversion rate prediction models corresponding to different marketing targets, as shown in model 1 and model 2 in Figure 7.

In a preferred embodiment 2, the advertisement delivery platform will match the attribute information of the current online user according to the targeting conditions selected when each advertisement is placed, and after matching, multiple advertisements will participate in the bidding, and the advertisement with the highest bid will be finalized Winning, after this step of screening, you can get the full set of advertisements to be exposed. For example: A, B, and C three advertisements, A chooses the orientation: Beijing, male, B chooses the orientation: Beijing, female, C chooses the orientation: Beijing, when a girl in Beijing visits Weibo, Only B and C meet the targeting conditions, so B and C enter the subsequent bidding process, and B and C will decide the winner according to the bidding rules.

score=m*bid

In the formula, score—the final bid price, m—the modified bid coefficient, bid—the default delivery price when the ad is launched.

In order to meet different marketing objectives, while referring to the preset delivery price at the time of advertising, and according to the target conversion rate of the marketing target selected at the time of advertising in the current traffic, a reward or penalty coefficient is made for the score, that is, the modified The bidding coefficient of , the calculation process is as follows:

According to the marketing target selected by the advertiser when placing the advertisement, select the target conversion rate model corresponding to the marketing target, and estimate the target conversion rate otr on the current traffic, which is specifically realized by the following formula:

otr=f(θ,x)

In the formula, f represents the estimated model function set by the advertising platform, θ represents the target conversion rate prediction model, and x represents the current traffic;

Please refer to FIG. 8, which is a schematic diagram of obtaining a target conversion rate reference value according to an embodiment of the present invention;

Store otr in the context of the advertisement, calculate the otr within the set second fixed time period as a reference value, and count it as otr', the value of the set second fixed time period is less than the set first fixed time period, Usually, the time period of 3 to 5 hours before the current time is selected, which is specifically realized by the following formula:

otr'=f(θ,x')

In the formula, f() represents the estimation model function set by the advertising delivery platform, θ represents the target conversion rate prediction model, and x' represents the traffic within the second fixed time period set before the current time.

Otr and otr’ have been optimized with a marketing target function of O, and the calculated bidding coefficient o(otr) is a floating-point number around 1.0. If it is greater than 1, it means that the current traffic is better than the historical traffic, otherwise it means that the traffic is worse than the historical traffic. Specifically, it is realized through the following formula:

In the formula: o() represents the marketing target optimization function preset by the advertising delivery platform, otr represents the target conversion rate, otr’ represents the reference value of the target conversion rate, and β represents the manual intervention coefficient.

Limit the value range of the bidding coefficient to prevent the maximum and minimum value marketing system revenue, and the revised bidding coefficient score is used for system ranking, and the winner gets exposure opportunities.

score=bid·Range{O(otr,otr'))

Range{O(otr,otr')}∈(0.65,1.35)

m=Range{O(otr,otr')}

In the formula, Range()-value function.

The embodiment of the present invention provides a system for optimizing and placing advertisements based on advertisement creativity, which can implement the method embodiments provided above. For specific function realization, please refer to the description in the method embodiments, and details will not be repeated here.

It is understood that the specific order or hierarchy of steps in the processes disclosed is an example of exemplary approaches. Based upon design preferences, it is understood that the specific order or hierarchy of steps in the processes may be rearranged without departing from the scope of the present disclosure. The accompanying method claims present elements of the various steps in a sample order, and are not meant to be limited to the specific order or hierarchy described.

In the foregoing Detailed Description, various features are grouped together in a single embodiment to simplify the disclosure. This method of disclosure is not to be interpreted as reflecting an intention that the claimed embodiments of the subject matter require more features than are expressly recited in each claim. Rather, as the following claims reflect, the invention lies in less than all features of a single disclosed embodiment. Thus the following claims are hereby expressly incorporated into the Detailed Description, with each claim standing on its own as a separate preferred embodiment of this invention.

The foregoing description of the disclosed embodiments was provided to enable any person skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may also be applied to other embodiments without departing from the spirit and scope of the disclosure. Thus, the present disclosure is not intended to be limited to the embodiments presented herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

The foregoing description includes illustrations of one or more embodiments. Of course, it is impossible to describe all possible combinations of components or methods to describe the above-mentioned embodiments, but those skilled in the art should recognize that various embodiments can be further combined and permuted. Accordingly, the embodiments described herein are intended to embrace all such alterations, modifications and variations that fall within the scope of the appended claims. Furthermore, to the extent that the term "comprises" is used in the specification or claims, the word is encompassed in a manner similar to the term "comprises" as interpreted when "comprises" is used as a link in the claims. Furthermore, any use of the term "or" in the specification of the claims is intended to mean a "non-exclusive or".

Those skilled in the art can also understand that various illustrative logical blocks, units, and steps listed in the embodiments of the present invention can be implemented by electronic hardware, computer software, or a combination of both. To clearly demonstrate the interchangeability of hardware and software, the various illustrative components, units and steps above have generally described their functions. Whether such functions are implemented by hardware or software depends on the specific application and overall system design requirements. Those skilled in the art may use various methods to implement the described functions for each specific application, but such implementation should not be understood as exceeding the protection scope of the embodiments of the present invention.

Various illustrative logic blocks or units described in the embodiments of the present invention can be discretely processed by a general-purpose processor, a digital signal processor, an application-specific integrated circuit (ASIC), a field programmable gate array or other programmable logic devices. Gate or transistor logic, discrete hardware components, or any combination of the above designed to implement or operate the described functions. The general-purpose processor may be a microprocessor, and optionally, the general-purpose processor may also be any conventional processor, controller, microcontroller or state machine. A processor may also be implemented by a combination of computing devices, such as a digital signal processor and a microprocessor, multiple microprocessors, one or more microprocessors combined with a digital signal processor core, or any other similar configuration to accomplish.

The steps of the method or algorithm described in the embodiments of the present invention may be directly embedded in hardware, a software module executed by a processor, or a combination of both. The software modules may be stored in RAM memory, flash memory, ROM memory, EPROM memory, EEPROM memory, registers, hard disk, removable disk, CD-ROM or any other storage medium in the art. Exemplarily, the storage medium can be connected to the processor, so that the processor can read information from the storage medium, and can write information to the storage medium. Optionally, the storage medium can also be integrated into the processor. The processor and the storage medium can be set in the ASIC, and the ASIC can be set in the user terminal. Optionally, the processor and the storage medium may also be set in different components in the user terminal.

In one or more exemplary designs, the above functions described in the embodiments of the present invention may be implemented in hardware, software, firmware or any combination of the three. If implemented in software, the functions can be stored on or transmitted over as one or more instructions or code on a computer-readable medium. Computer-readable media includes computer storage media and communication media that facilitate transfer of a computer program from one place to another. Storage media may be any available media that can be accessed by a general purpose or special computer. For example, such computer-readable media may include, but are not limited to, RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage or other magnetic storage devices, or any other device that can be used to carry or store instructions or data structures and Other medium of program code in a form readable by a general-purpose or special-purpose computer, or a general-purpose or special-purpose processor. In addition, any connection is properly defined as a computer-readable medium, for example, if the software is transmitted from a website site, server, or other remote source via a coaxial cable, fiber optic cable, twisted pair, digital subscriber line (DSL) Or transmitted by wireless means such as infrared, wireless and microwave are also included in the definition of computer readable media. Disks and discs include compact discs, laser discs, optical discs, DVDs, floppy discs, and Blu-ray discs. Disks usually reproduce data magnetically, while discs usually reproduce data optically with lasers. Combinations of the above can also be contained on a computer readable medium.

The specific embodiments described above have further described the purpose, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above descriptions are only specific embodiments of the present invention and are not intended to limit the scope of the present invention. Protection scope, within the spirit and principles of the present invention, any modification, equivalent replacement, improvement, etc., shall be included in the protection scope of the present invention.

Claims (10)

1. A method for optimizing advertisement delivery based on advertising creativity, characterized in that the method comprises: Preset the corresponding relationship between the marketing objectives provided by the advertising delivery platform and the user behavior on the advertising delivery platform; According to the preset corresponding relationship and different marketing goals, the user's historical behavior on the advertising platform is used to perform prediction model training to obtain a target conversion rate prediction model corresponding to the marketing goal. Historic user behaviors include: all user behaviors on the advertising delivery platform during the first fixed time period set before the current time; Calculate the bidding coefficient of the advertisement according to the marketing goal selected when the advertisement is launched, and the target conversion rate estimation model corresponding to the marketing goal; According to the bidding coefficient of each advertisement and the preset bidding rules, the final bid of each advertisement is obtained, and the winning advertisement is exposed to users who meet the targeting conditions selected when the advertisement is placed. 2. The method for optimizing advertisement placement based on advertisement creativity according to claim 1, wherein the method further comprises: For the current user of the advertisement delivery platform, match according to the targeting condition selected when the advertisement is placed, and filter out the complete set of advertisements to be exposed. The targeting condition selected during the advertisement delivery is the attribute information of the pre-advertising user; Calculate the bidding coefficient of each advertisement according to the marketing target selected when each advertisement in the full set of advertisements to be exposed is placed, and the target conversion rate estimation model corresponding to the marketing target; According to the bidding coefficient of each advertisement and the preset bidding rules, the final bid of each advertisement in the full set of advertisements to be exposed is obtained, and the winning advertisement is delivered to the current user. 3. The method for optimizing advertisement placement based on advertisement creativity according to claim 1 or 2, wherein the marketing objectives provided by the preset advertisement placement platform include: increasing the number of fans, increasing customer consultation, and increasing brand awareness; The user behavior on the advertising delivery platform includes: following, forwarding, commenting, liking, saving, clicking on the profile picture, and clicking on the details page; The corresponding relationship between the marketing goals provided by the preset advertising delivery platform and user behavior on the advertising delivery platform includes: increasing the number of fans corresponding to attention behaviors, increasing customer consultation and comments, and corresponding behaviors of clicking details pages, increasing brand awareness Corresponding to following, reposting, commenting, liking, bookmarking, clicking on profile pictures, and clicking on details pages; According to the preset corresponding relationship and different marketing objectives, the estimation model training is carried out with the historical behavior of users on the advertising delivery platform, and the target conversion rate estimation model corresponding to the marketing objectives is obtained, which specifically includes: Collect all user behaviors on the advertising delivery platform within the set first fixed time period, and use the presence or absence of user behaviors corresponding to different preset marketing objectives as training samples; A known click through rate prediction model is trained offline according to the training samples to obtain a target conversion rate prediction model corresponding to the marketing target. 4. The method for optimizing advertisement placement based on advertisement creativity according to claim 1 or 2, characterized in that the calculation method of the bidding coefficient of the advertisement specifically includes: According to the preset marketing target when the advertisement is launched, the target conversion rate on the current traffic is estimated by using the target conversion rate estimation model corresponding to the marketing target; Store the target conversion rate on the current traffic in the context of the corresponding advertisement, and calculate the reference value of the target conversion rate in the second fixed time period set before the current time, and the value of the second fixed time period is less than the value of the first fixed time period; According to the marketing target optimization function preset by the advertising delivery platform, as well as the target conversion rate and the reference value of the target conversion rate, the bidding coefficient of each advertisement is calculated. 5. The method for optimizing advertisement placement based on advertisement creativity according to claim 1 or 2, wherein the winning advertisement refers to the advertisement with the highest final bid; The calculation method of the final bid of the advertisement, specifically including: According to the value function, the value of the bidding coefficient of the advertisement is controlled within a predetermined value range, and the revised bidding coefficient is obtained; The product of the preset delivery price of the ad and the revised bidding coefficient is used as the final bid of the ad. 6. A system for optimizing advertising based on advertising creativity, characterized in that the device includes: The preset unit is used to preset the corresponding relationship between the marketing objectives provided by the advertising delivery platform and the user behavior on the advertising delivery platform; The training unit is used to perform estimation model training based on the preset corresponding relationship and different marketing objectives based on the historical behavior of users on the advertising delivery platform to obtain a target conversion rate estimation model corresponding to the marketing objectives. The historical behavior of users on the advertising delivery platform includes: the behavior of all users on the advertising delivery platform mentioned in the first fixed time period set before the current time; The first calculation unit is used to calculate the bidding coefficient of the advertisement according to the marketing target selected when the advertisement is launched, and the target conversion rate estimation model corresponding to the marketing target; The first bidding unit is used to obtain the final bid of each advertisement according to the bidding coefficient of each advertisement and the preset bidding rules, and expose the winning advertisement to users who meet the targeting conditions selected when the advertisement is delivered. 7. The system for advertising based on advertising creative optimization according to claim 6, characterized in that the device further comprises: The screening unit is used to match the current user of the advertisement delivery platform according to the targeting condition selected when the advertisement is delivered, and filter out the complete set of advertisements to be exposed, and the targeting condition selected during the advertisement delivery is the attribute information of the pre-advertising user; The first calculation unit is also used to calculate the bidding coefficient of each advertisement according to the marketing target selected when each advertisement in the full set of advertisements to be exposed is placed, and the target conversion rate estimation model corresponding to the marketing target; The first bidding unit is further used to obtain the final bid of each advertisement in the full set of advertisements to be exposed according to the advertisement bidding coefficient and preset bidding rules, and deliver the winning advertisement to the current user. 8. The system for advertising based on advertising creativity optimization according to claim 6 or 7, characterized in that, In the preset unit, the marketing goals provided by the preset advertising delivery platform include: increasing the number of fans, increasing customer consultation, and increasing brand awareness; In the preset unit, user behaviors on the advertising delivery platform include: follow, forward, comment, like, bookmark, click on profile picture, click on details page; In the preset unit, the correspondence between the marketing objectives provided by the preset advertising platform and user behavior on the advertising platform includes: increasing the number of fans corresponding to user attention behavior, increasing customer consultation and user comments, and clicking on the details page Behavioral correspondence, increasing brand awareness corresponds to user attention, forwarding, commenting, likes, favorites, clicking on avatars, and clicking on details pages; The training unit specifically includes: The collection module is used to collect all user behaviors on the advertising delivery platform within the set first fixed time period, and obtain the presence or absence of user behaviors corresponding to different preset marketing objectives as training samples; The training module is used to train a known click-through rate prediction model offline according to the training samples, so as to obtain a target conversion rate prediction model corresponding to the marketing target. 9. The system according to claim 6 or 7, wherein the advertisement delivery system based on advertisement creative optimization is characterized in that, in the first calculation unit, the module for calculating the advertisement bidding coefficient specifically includes: The first calculation sub-module is used for estimating the target conversion rate on the current traffic by using the target conversion rate estimation model corresponding to the marketing target according to the preset marketing target when the advertisement is placed; The second calculation sub-module is used to store the target conversion rate on the current traffic in the context of the corresponding advertisement, and calculate the reference value of the target conversion rate in the second fixed time period set before the current time, said The value of the second fixed time period is less than the value of the first fixed time period; The third calculation sub-module is used to calculate the bidding coefficient of each advertisement according to the marketing target optimization function preset by the advertising delivery platform, as well as the target conversion rate and the reference value of the target conversion rate. 10. The system for optimizing delivery targets based on advertising creativity according to claim 6 or 7, wherein the winning advertisement refers to the advertisement with the highest final bid; The first bidding unit includes a calculation module for the final bid of the advertisement, and the calculation module specifically includes: The value-taking sub-module is used to control the value of the bidding coefficient of the advertisement within a predetermined value range according to the value-taking function, so as to obtain the revised bidding coefficient; The scoring sub-module is used to use the product of the preset delivery price of the advertisement and the revised bidding coefficient as the final bidding price of the advertisement.