CN114862464A

CN114862464A - Advertisement putting effect estimation method and device

Info

Publication number: CN114862464A
Application number: CN202210499575.7A
Authority: CN
Inventors: 陈珺; 张泽晗; 黄思光; 喻川
Original assignee: Alibaba China Co Ltd
Current assignee: Alibaba China Co Ltd
Priority date: 2022-05-09
Filing date: 2022-05-09
Publication date: 2022-08-05

Abstract

The embodiment of the specification provides an advertisement putting effect estimation method and an advertisement putting effect estimation device, wherein the advertisement putting effect estimation method comprises the following steps: acquiring historical flow and advertisement putting information, wherein the advertisement putting information comprises bidding information; determining initial estimated flow according to the advertisement putting information and the historical flow; determining target estimated flow in the initial estimated flow according to the bidding information; and counting estimated delivery effect information of the advertisement delivery information based on the target estimated flow. The initial estimated flow is determined through the advertisement delivery information and the historical flow, and the target estimated flow is determined in the initial estimated flow, so that the estimated delivery effect information of the advertisement delivery information can be counted according to the target estimated flow and the bidding information, the estimated delivery effect information of the advertisement delivery information is provided for a user, the insight is provided for the user, the advertisement delivery information can be optimized in advance, and the trial and error cost is reduced to the maximum extent.

Description

Advertisement putting effect estimation method and device

Technical Field

The embodiment of the specification relates to the technical field of advertisement putting, in particular to an advertisement putting effect estimation method. One or more embodiments of the present disclosure also relate to an advertisement delivery effect prediction apparatus, a computing device, and a computer-readable storage medium.

Background

At present, when an advertiser sets an advertisement putting plan on an advertisement platform for marketing promotion, the advertiser usually observes the putting effect of the advertisement after putting the advertisement putting plan for a period of time, and then adjusts and optimizes the advertisement putting plan based on own putting experience. In the adjusting and optimizing process, the advertiser inevitably generates some trial and error costs, and wastes a large amount of manpower and budget, so when the advertiser sets an advertisement delivery plan, how to provide the advertiser with an advertisement delivery effect for a period of time in the future in real time to reduce the cost of advertisement delivery of the advertiser is a problem which needs to be solved urgently at present.

Disclosure of Invention

In view of this, the embodiments of the present specification provide an advertisement delivery effect estimation method. One or more embodiments of the present disclosure also relate to an advertisement delivery effect estimation apparatus, a computing device, a computer-readable storage medium, and a computer program, so as to solve the technical defects in the prior art.

According to a first aspect of embodiments of the present specification, there is provided an advertisement placement effect estimation method, including:

acquiring historical flow and advertisement putting information, wherein the advertisement putting information comprises bidding information;

determining initial estimated flow according to the advertisement putting information and the historical flow;

determining target estimated flow in the initial estimated flow according to the bidding information;

and counting estimated delivery effect information of the advertisement delivery information based on the target estimated flow.

According to a second aspect of embodiments of the present specification, there is provided an advertisement placement effect prediction apparatus including:

the system comprises an acquisition module, a storage module and a display module, wherein the acquisition module is configured to acquire historical traffic and advertisement delivery information, and the advertisement delivery information comprises bidding information;

a first determination module configured to determine an initial estimated traffic according to the advertisement delivery information and the historical traffic;

a second determination module configured to determine a target pre-estimated flow rate from the initial pre-estimated flow rates according to the bidding information;

and the statistic module is configured to count estimated delivery effect information of the advertisement delivery information based on the target estimated flow.

According to a third aspect of embodiments herein, there is provided a computing device comprising a memory, a processor and computer instructions stored on the memory and executable on the processor, the processor implementing the steps of the advertisement impression prediction method when executing the computer instructions.

According to a fourth aspect of the embodiments of the present specification, there is provided a computer-readable storage medium storing computer instructions which, when executed by a processor, implement the steps of the advertisement impression effect estimation method.

According to a fifth aspect of embodiments of the present specification, there is provided a computer program, wherein when the computer program is executed in a computer, the computer is caused to execute the steps of the advertisement impression effect estimation method.

The advertisement delivery effect estimation method provided by the specification acquires historical flow and advertisement delivery information, wherein the advertisement delivery information comprises bidding information; determining initial estimated flow according to the advertisement putting information and the historical flow; determining target estimated flow in the initial estimated flow according to the bidding information; and counting estimated delivery effect information of the advertisement delivery information based on the target estimated flow.

The embodiment of the specification realizes that the initial estimated flow is determined through the advertisement delivery information and the historical flow, and the target estimated flow is determined in the initial estimated flow, so that the estimated delivery effect information of the advertisement delivery information can be counted according to the target estimated flow and the bidding information, the estimated delivery effect information of the advertisement delivery information is provided for a user when the user takes the advertisement delivery information as an advertisement delivery plan, the insight is provided for the user, the advertisement delivery information can be optimized in advance, and the trial-and-error cost is reduced to the maximum extent.

Drawings

Fig. 1 is a flowchart of an advertisement delivery effect estimation method according to an embodiment of the present specification;

FIG. 2 is a schematic diagram of a system for advertisement placement effectiveness estimation according to an embodiment of the present disclosure;

fig. 3 is a schematic diagram of a technical architecture of an offline RTP model provided in an embodiment of the present specification;

fig. 4 is a flowchart illustrating a processing procedure of a method for estimating advertisement delivery effectiveness according to an embodiment of the present disclosure;

fig. 5 is a schematic structural diagram of an advertisement delivery effect estimation apparatus according to an embodiment of the present disclosure;

fig. 6 is a block diagram of a computing device according to an embodiment of the present disclosure.

Detailed Description

In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present description. This description may be embodied in many different forms and should not be construed as limited to the embodiments set forth herein, as those skilled in the art will be able to make and use the present disclosure without departing from the spirit and scope of the present disclosure.

The terminology used in the description of the one or more embodiments is for the purpose of describing the particular embodiments only and is not intended to be limiting of the description of the one or more embodiments. As used in one or more embodiments of the present specification and the appended claims, the singular forms "a," "an," and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It should also be understood that the term "and/or" as used in one or more embodiments of the present specification is intended to encompass any and all possible combinations of one or more of the associated listed items.

It will be understood that, although the terms first, second, etc. may be used herein in one or more embodiments to describe various information, these information should not be limited by these terms. These terms are only used to distinguish one type of information from another. For example, a first can also be referred to as a second and, similarly, a second can also be referred to as a first without departing from the scope of one or more embodiments of the present description. The word "if" as used herein may be interpreted as "at … …" or "when … …" or "in response to a determination", depending on the context.

First, the noun terms to which one or more embodiments of the present specification relate are explained.

And (3) advertisement putting plan: the carriers for advertisers to put advertisements generally comprise elements such as bidding price, budget, targeted population, promoted goods, originality and the like;

putting in effect: and (4) the effect of advertisement delivery within a preset time, such as the number of covered crowds, the display amount, the click amount, the consumption and the like.

Estimating the probability: commonly referred to as estimates, estimated probabilities of certain behavior occurring on the flow (consumer), such as click-through rates, purchase collection rates, etc.

An advertisement system: the system link for realizing advertisement delivery generally comprises links such as recall, estimation, sequencing, strategy, fee deduction and the like.

Log playback: and reproducing the bidding process and bidding results of the historical flow on a landing log of the official port system.

CPC: cost-per-click, CPC refers to click billing, which is based on the price per advertisement click. For example, if the price of a certain advertisement per click is 0.5 yuan, the CPC is 0.5.

CPM: cost Per mile, CPM, is the Cost that an advertisement needs to spend for every thousand people presented, and so is the Cost of thousands of people presented. For example, if the price of a thousand exposures for a certain ad slot is 10 yuan, the CPM is 10.

BCB: the Budget Constrained Bidding, BCB is the automatic Bidding of the advertising platform under the Budget constraint, i.e. the advertising platform has possibly more auction traffic under the highest Budget given by the advertiser.

MCB: Multi-Constrained Bidding, MCB offers the same idea as BCB, but MCB may place other constraints besides constraints on budget, e.g., the advertiser wants the click cost not to be too high.

Currently, an advertisement platform provides online advertisement service for merchants, and assists advertisers to launch advertisements online, and advertisers set an advertisement launching plan on the advertisement platform for marketing promotion, and they usually observe the launching effect of the advertisements after the advertisement launching plan is launched for a period of time, for example, observe the launching effect of the advertisements after the advertisement launching plan is implemented for one day. Then, the advertiser can optimize and adjust the advertisement putting plan based on the putting effect and the own putting experience so as to achieve better advertisement putting effect. In this process, the advertiser inevitably generates a series of trial and error costs, which wastes a lot of manpower and budget costs, and especially when the advertiser has no delivery experience, the trial and error costs are higher. The advertising platform has the significance of simplifying and centralizing the setting, execution and optimization process of the plan, and helping to improve the operation efficiency of the client. Obviously, when the advertiser sets the advertisement delivery plan, if the platform can provide the delivery effect of the future day in real time to be displayed to the client, such as the audience number, the display number, the click number and the like, the platform can provide insight for the advertiser client, so that the advertiser client can make optimization decisions in advance, the trial and error cost is reduced to the greatest extent, the stickiness of the advertiser client is enhanced, and more platform budgets are attracted.

Based on this, in the present specification, an advertisement delivery effect estimation method is provided for providing an estimated delivery effect corresponding to an advertisement delivery plan for an advertiser client, and the present specification also relates to an advertisement delivery effect estimation apparatus, a computing device, a computer-readable storage medium, and a computer program, which are described in detail one by one in the following embodiments.

Fig. 1 shows a flowchart of an advertisement placement effectiveness estimation method provided in accordance with an embodiment of the present specification, which includes steps 102 to 108.

Currently, real-time bidding (RTB) is a popular mode in online advertising, advertisers can bid on advertisement presentation opportunities at a traffic level and can flexibly select when, where, and to whom the advertisements are presented, and the combination of these settings made by advertisers is also referred to as an advertisement delivery plan or advertisement delivery information. Since an advertiser needs to observe a corresponding effect after a certain time of delivery in an advertisement delivery plan, a large amount of manpower and material resources may be wasted, so an estimation method suitable for a presentation billing (CPM) bid plan is proposed in the industry, under the current advertisement delivery plan setting, an advertisement system bidding process of the current advertisement delivery plan on historical traffic is played back, a current plan competition traffic set is obtained, in the competition traffic set, delivery success (presentation amount, consumption cost and the like) is counted and planned based on the traffic amount and market price, but the method lacks historical traffic estimation probability information of the current advertisement delivery plan, and cannot provide estimation for other advertisement bid types, such as CPC, MCB, BCB and the like.

In summary, the estimation of advertisement delivery effect needs to be applicable to various advertisement bid types, such as the manual bid CPM and CPC and the intelligent bid BCB and MCB, and can return accurate estimation effect for the advertiser in second-class time. The advertisement putting effect estimation method provided by the specification is used for achieving the target.

Step 102: obtaining historical traffic and advertisement delivery information, wherein the advertisement delivery information comprises bidding information.

The historical traffic can be understood as the access volume and click volume of all users in the past period of time. The advertisement delivery information may be understood as an advertisement delivery plan set by an advertiser, and the advertisement delivery information may include bidding information, user tag information, commodity tag information, and the like. The bid information may be understood as an advertisement bid type and a budget set by the advertiser in the advertisement delivery plan, for example, the advertisement bid type is CPC, and the budget is 0.5; or the advertisement bid type is CPM, and the budget is 50.

In practical application, the advertisement platform stores daily historical traffic, and in order to estimate the information of the advertisement effect of the current advertisement delivery information, it is necessary to determine the corresponding available traffic from the historical traffic according to the current advertisement delivery information, that is, if the current advertisement delivery information is delivered in the past, the traffic information such as the browsing volume, the click volume, the purchase volume, etc. of the user can be obtained in the past period of time.

In an embodiment of the present specification, advertisement delivery information set by an advertiser and historical traffic in a past day are obtained, the advertisement delivery information includes bid information, the bid information is a CPC advertisement bid type, and a CPC is 1.

Step 104: and determining initial estimated flow according to the advertisement putting information and the historical flow.

The initial estimated traffic can be understood as the traffic corresponding to the audience user determined in the historical traffic, that is, all estimated traffic that can be obtained according to the advertisement delivery information without considering the bid.

In practical application, the historical traffic corresponds to a plurality of users, but not all users are interested in the advertisements delivered by the advertisement delivery information, so that the audience user set and the initial estimated traffic corresponding to the audience user set can be screened out according to the advertisement delivery information.

Specifically, the advertisement delivery information includes user tag information;

determining initial estimated flow according to the advertisement putting information and the historical flow, wherein the method comprises the following steps:

determining an audience user set according to the user tag information;

and determining initial estimated flow according to the historical flow and the audience user set.

For example, if the advertisement put by the advertiser this time is female-oriented, it can be determined that each user in the audience user set is a female user, and the advertisement putting effect can be improved by determining the audience user set, that is, potential users interested in the advertisement this time are selected. The user label information can be understood as the characteristic information of each user, the user label information can be the information of occupation, years, purchase records and the like of the user, and whether the corresponding user is a potential audience user who puts the advertisement at this time can be determined according to the user label information. Determining the audience user set according to traffic may be understood as a recall task in the advertising system, i.e., determining which user groups the advertisement needs to be exposed to.

In practical application, the advertisement platform can obtain user attribute information of each user, the user attribute information includes occupation, years and other information corresponding to each user, the user attribute information includes information filled by the user when registering an account, and the information marked by the platform according to the historical behavior of the user. According to the comparison between the user label information in the advertisement putting information and the user attribute information of each user, the audience user set is determined, and the click rate, the conversion rate and other information of each user on the advertisement can be estimated subsequently.

In an embodiment of the present specification, user tag information in advertisement delivery information is obtained, and if the user tag information is an university student under 20 years old, an audience user set is determined according to the user tag information, each user in the audience user set conforms to the user tag information, and then initial estimated traffic corresponding to the advertisement delivery information of this time is determined according to the audience user set and historical traffic.

Specifically, determining an audience user set according to the user tag information includes:

acquiring a historical user set corresponding to the historical traffic;

screening in the historical user set according to user tag information in the advertisement putting information;

and determining an audience user set according to the screening result.

The historical user set can be understood as a user set formed by each corresponding historical user in the historical traffic. Each historical user can generate a plurality of flows, and a historical user set can be determined according to the historical flows, wherein the historical user set is a set of all users generating all the historical flows.

In practical application, when one user clicks to browse a commodity once, one flow is generated, and when the same user browses the same commodity for multiple times, multiple flows are generated; the historical user set corresponding to the historical traffic can be obtained according to the browsing behavior of each user, and the audience user set is subsequently screened.

In an embodiment of the present specification, a corresponding historical user set is determined according to historical traffic, the historical user set includes 3000 people, screening is performed in the historical user set according to user tag information in advertisement delivery information, a screening result is 1000 people, then 1000 people are audience users, each user is characterized to be in accordance with the user tag information, and the 1000 people are formed into an audience user set.

Specifically, the advertisement delivery information includes product label information;

determining initial estimated flow according to the historical flow and the audience user set, wherein the method comprises the following steps:

determining the flow information of each audience user in the audience user set according to the historical flow;

determining the scoring information of each audience user in the audience user set according to the commodity label information in the advertisement putting information;

and determining initial estimated flow according to the flow information and the grading information of each audience user.

The traffic information of each audience user can be understood as the traffic generated by each audience user, the score information can be understood as the score of each audience user corresponding to the advertisement put in this time, the score information can be information such as the click rate and the conversion rate of each audience user to the advertisement put in this time, the higher the score information is, the higher the interest of the user to the advertisement put in this time is, and the user score information is needed to be calculated when the putting effect information of the advertisement putting information is subsequently calculated. The product label information can be understood as attribute information of a product displayed in the advertisement released at this time, the product label information can include information such as a product picture and product introduction, and score information of each audience user can be estimated according to the product label information, that is, score information of flow rate of each audience user by an advertiser represents whether the flow rate can be scored by clicking, purchasing and the like.

In practical application, after an advertiser sets advertisement putting information, the initial estimated flow with the personalized estimated score can be determined in time, and through a bidding link, historical auction flow with the personalized estimated score is determined, so that a good data base is provided for subsequent calculation of putting effect information.

In an embodiment of the description, browsing information of 1000 audience users in an audience user set is determined according to historical traffic, each audience user is scored according to commodity label information in advertisement delivery information and browsing information of each audience user, and initial estimated traffic with personalized estimated scores is determined according to the browsing information of each audience user and corresponding scoring information.

Step 106: and determining target estimated flow in the initial estimated flow according to the bidding information.

The bidding information can be understood as budget and bidding type information set in the advertisement putting information by an advertiser, the bidding information includes the budget of the advertisement putting and the bidding type of the advertisement putting, and in different bidding types, different bidding conditions are also defined in the bidding information. For example, in the CPC manual bid type, the price of each click is defined in the bidding information, and then the price and budget of each click need to be considered when a subsequent flow auction is performed; under the BCB automatic bidding type, budget is defined in bidding information, and only the budget needs to be considered when a subsequent flow auction is carried out. The target estimated traffic can be understood as the traffic corresponding to the audience user determined in the initial estimated traffic, that is, all estimated traffic that can be obtained under the condition of considering the bid according to the advertisement delivery information.

In practical application, in the determined initial estimated traffic, because different traffic may come from different advertisement showing slots, the price of each advertisement showing slot is different, and the bid types set by each schedule may also be different, the target estimated traffic is the budget and bid types considered by the advertiser for this time consumption. Under the same budget, different bid types may also obtain different traffic flows, so that the target estimated traffic flow needs to be determined from the initial estimated traffic flow according to the bid information in the advertisement delivery information.

In an embodiment of the present specification, a flow auction is performed from the initial estimated flow according to the budget cost and the manual bid type in the bidding information, and the target estimated flow is determined according to the auction result.

Specifically, determining the target estimated traffic in the initial estimated traffic according to the bidding information in the advertisement delivery information includes:

determining a bidding type according to bidding information in the advertisement putting information;

under the condition that the bidding type is a first bidding type, determining target estimated traffic in the initial estimated traffic according to unit price information in the bidding information;

and under the condition that the bidding type is a second bidding type, determining the target estimated flow in the initial estimated flow according to the total price information in the bidding information.

The bid type can be understood as an advertisement bid type set by an advertiser in the advertisement delivery information, and the bid type can include a manual bid type, such as a CPC bid type, a CPM bid type, and the like; and may also be an automatic bid type such as a BCB bid type, an MCB bid type, and the like. And under different bid types, the logic for subsequently determining the target estimated flow is different. The first bid type can be understood as a manual bid type, under the manual bid type, an advertiser can give a fixed bid to participate in subsequent flow bidding, and the fixed bid given by the advertiser is the single-price information in the bidding information; the second bid type may be understood as an automatic bid type in which an advertiser only gives a fixed budget and wishes to obtain as much of the total value of the presentation as possible under a limited budget.

In practical application, because we need to offer in a plurality of different bid typesWhen the auction of the target estimated traffic is performed under the initial estimated traffic, the representation needs to support the traffic auction under both the manual bid type (CPC, CPM, etc.) and the automatic bid type (BCB, MCB, etc.), and in general, a second price auction mechanism is used to simulate the bidding process of the advertiser when participating in the auction. The second price auction mechanism can be understood as a mechanism for making an auction result at the second highest price during a real-time auction (RTB). Specifically, assuming that N display slots are ordered in index i order during a day, the advertiser places a bid b for traffic i _i And compete with other bidders in real time. If b is _i Is the highest bid in this auction, the advertiser will bid at c _i The price of (the second highest price), which is the highest bid by the other competitors, wins the show opportunity i. The bidding process is terminated when the delivery result reaches the limit set by the advertiser, for example, the total consumption reaches the preset budget and the auction of the exhibition opportunity is completed. Note that because multiple plans of different bid types compete for the same presentation opportunity, b herein _i Has been converted to universal currency expressions.

For the manual bid type, the advertiser offers a fixed bid to participate in a traffic bid, which corresponds to a monetization b in the plan to eventually win the traffic _i Is the highest bid, manual bidding can control costs because the price charged per winning flow is lower than b _i ，b _i The bid price information is calculated based on bid price information in advertisement putting information set by an advertiser, and the calculation formula is as follows, wherein bid is the bid price information in the bid price information, pctr _i Is the estimated click rate of the flow.

In an embodiment of the present specification, a bidding type is determined as a manual bidding CPM type according to bidding information in advertisement delivery information, and then bidding is performed in an initial estimated traffic according to unit price information in the bidding information to obtain a finally determined target estimated traffic.

In practical applications, for the automatic bidding (BCB, MCB) type, the advertiser wants to obtain as much total value of the show (BCB) as possible under a limited budget, and the cost of unit performance acquisition does not exceed the preset constraint (MCB), assuming x _i Representing whether the advertiser wins a show opportunity i, v _i Representing the value of traffic to the advertiser, the BCB formula definition is shown in equation 2, and the MCB formula definition is shown in equation 3:

max x _i *v _i ,BCB

s.t.∑x _i *c _i equation 2

max x _i *v _i ，MCB

s.t.∑x _i *c _i ＜＝budget

In formula 2 and formula 3, the budget may be understood as a budget in the advertisement delivery information set by the advertiser, and the constraint may be understood as a preset cost for acquiring unit performance. v. of _i Is a value expression of the effect of the advertisement after the advertisement is put, which is set by the advertiser.

In an embodiment of the present specification, the bidding type is determined to be an automatic bidding BCB type according to bidding information in the advertisement delivery information, and then bidding is performed in the initial estimated traffic according to total price information budget in the bidding information to obtain the finally determined target estimated traffic.

Specifically, determining the target estimated traffic in the initial estimated traffic according to the unit price information in the bidding information includes:

comparing unit price information in the bidding information with historical unit price information in the flow bidding of the initial estimated flow;

under the condition that the comparison result meets a unit price bidding condition, acquiring the estimated flow corresponding to the comparison result;

and determining the target estimated flow according to the estimated flow corresponding to the at least one round of comparison result.

The historical unit price information can be understood as historical prices made by other competitors in flow bidding of the initial estimated flow, and by comparing the unit price information in the current bidding information with the historical unit price information, which estimated flows can be obtained through bidding auction according to the current advertisement putting information can be determined, so that the target estimated flow is determined. The unit price bidding condition can be understood as that when the unit price bidding condition is met, the bidding succeeds and the corresponding estimated flow is obtained, wherein the unit price bidding condition comprises that the historical unit price information is less than or equal to the unit price information, namely, the flow with the historical unit price information less than or equal to the unit price information of this time is selected from the initial estimated flow.

In practical application, if there may be a plurality of display positions and a plurality of periods of flow in the initial estimated flow for bidding auction, comparing the historical unit price information in each bidding auction process with the unit price information in the current bidding information, so as to obtain the target estimated flow that can be obtained by the current advertisement delivery information.

In an embodiment of the present specification, in each round of bidding of the initial estimated flow, historical unit price information and unit price information in the current bidding information are compared, a flow in which the historical unit price information is less than or equal to the unit price information in the current bidding information is selected, and a target estimated flow is determined.

Specifically, determining the target estimated flow rate in the initial estimated flow rate according to the total price information in the bidding information includes:

comparing the total price information in the bidding information with historical total price information in the flow bidding of the initial estimated flow;

under the condition that the comparison result meets the total price bidding condition, acquiring the estimated flow corresponding to the comparison result;

The historical total price information can be understood as prices spent on obtaining a plurality of estimated flows in the initial estimated flows, the total price bidding condition can be understood as that the historical total price information is less than or equal to the total price information, and the total price bidding condition comprises that the historical total price information is less than or equal to the total price information.

In practical application, in the process of bidding in the automatic bidding type, we will determine an automatic bidding strategy, in which a bid for a show opportunity i is given _i Is based on the value v of the presentation opportunity _i Determining that the bidding form of the automatic bidding strategy is shown as formula 4:

where α, β, γ are scaling factors, constraint can be understood as the cost of a predetermined unit yield acquisition. After the automatic bidding is determined based on the automatic bidding strategy, historical total price information spent is determined based on the estimated flow obtained by the automatic bidding. The MCB bid type differs from the BCB bid type in that additional observation is required to see whether the unit cost reaches a preset constraint.

In one embodiment of the present specification, an auto bid is first determined according to an auto bid policy _i And selecting estimated flow with historical unit price information less than or equal to the automatic bidding, sequencing the determined estimated flow from low to high according to the historical unit price information, and accumulating the historical unit price information corresponding to each estimated flow from low to high in sequence to obtain historical total price information until the historical total price information reaches the total price information in the bidding information.

Step 108: and counting estimated delivery effect information of the advertisement delivery information based on the target estimated flow.

After the target estimated traffic obtained through the bidding auction is known, the estimated delivery effect information of the advertisement delivery information can be counted according to the target estimated traffic.

The estimated delivery effect information can be understood as the estimated delivery effect of the current advertisement delivery information, the estimated delivery effect information comprises the sum of estimated showing values of all estimated traffic, the sum of estimated click rate and the sum of estimated conversion rate, and the estimated delivery effect information can be understood as a triple, wherein the estimated showing value is a value in a range of 0 to 1, the larger the estimated showing value is, the advertisement is represented to the user to watch, and the estimated showing value is kept as 1 because the user is determined in the specification; the estimated click rate is the probability of the user clicking the advertisement put at this time, the estimated click rate is a value in the interval of 0% to 100%, and the higher the estimated click rate is, the more the user clicks the advertisement to browse; the estimated conversion rate is the estimated probability that the user performs other effective behaviors in the sum of browsing the advertisements, and the effective behaviors can include but are not limited to behaviors of paying, collecting and the like.

In practical application, because the target estimated traffic is known, estimated delivery effect information of each estimated traffic on the advertisement delivered at this time can be obtained, and the estimated delivery effect information of the advertisement delivery information at this time is obtained by accumulating the estimated delivery effect information.

In an embodiment of the present specification, unit price information of each estimated flow is determined according to a target estimated flow, an estimated click rate and an estimated conversion rate of each estimated flow are obtained based on score information of each estimated flow, a sum of the estimated conversion rates after counting unit price information of all estimated flows, after estimating the click rate, is calculated, and estimated delivery effect information of the current advertisement delivery information is obtained.

The target estimated traffic comprises score information of audience users corresponding to each estimated traffic;

the estimated delivery effect information of the advertisement delivery information is counted based on the target estimated flow, and the method comprises the following steps:

and according to each estimated flow and the grading information corresponding to each estimated flow, counting estimated delivery effect information of the advertisement delivery information.

The scoring information corresponding to each estimated flow can be understood as the scoring information of the user corresponding to the estimated flow, and the click rate and the conversion rate corresponding to each estimated flow can be estimated according to the scoring information of each estimated flow.

In an embodiment of the application, the estimated click rate of each estimated flow is determined according to the scoring information corresponding to each estimated flow, the price, click rate and exhibition value spent on obtaining the target estimated flow according to the advertisement delivery information are counted, the total consumption price, total click rate and total exhibition value in the estimated delivery effect information are determined according to the counting result, and the estimated delivery effect information is fed back to the advertiser.

In practical application, because the estimated delivery effect is obtained according to the score information statistics of the audience users corresponding to each estimated flow, an error necessarily exists between the estimated delivery effect and the real delivery effect, and therefore the estimated delivery effect can be calibrated, and the estimated delivery effect closer to the real delivery effect can be obtained, specifically, the method further comprises the following steps:

calculating calibration information corresponding to the estimated putting effect information according to the estimated putting effect information and the advertisement putting information;

and calibrating the estimated putting effect information according to the calibration information, and obtaining the calibrated estimated putting effect information.

The calibration information can be understood as a coefficient for calibrating the triple in the estimated delivery effect information, and the estimated delivery effect closer to the real delivery effect can be obtained by calibrating the triple in the estimated delivery effect information according to the calibration information.

In an embodiment of the present specification, a corresponding calibration coefficient is calculated and obtained according to a triplet in the estimated delivery effect information and advertisement delivery information, a calibrated triplet is calculated according to the calibration coefficient and a total showing value, a total click rate, and a total consumption price in the triplet, and the calibrated triplet is used as the estimated delivery effect information fed back to the user.

In practical application, the accuracy of the estimated delivery effect information can be measured by using an index that maps represent the error percentage between the estimated value and the true value, maps can be obtained by calculation according to formula (5), and ratio _0.5 Ratio of sample number, ratio, characterizing the error of estimated percentage within 50% _0.5 Can be obtained by calculation of formula (6).

Wherein, ape _i Representing the percentage of error between the predicted and actual values of a predicted flow, y' _i Can be understood as an estimate, y _i As can be appreciated, true, map is the average error percentage for all predicted flows,

it should be noted that the percentage error of the ratio samples can be determined according to actual conditions, for example, if the ratio samples with the estimated percentage error within 60% need to be calculated, the ratio corresponding to the ratio samples with the estimated percentage error within 60% can be calculated _0.6 . In practical application, the measure is calculated as map 0.63, ratio _0.5 This solution brings about a profit of the customer ARPU + 3%.

In practical application, the advertisement delivery effect estimation method provided by the present specification may also be implemented by using a trained model, specifically, referring to fig. 2, fig. 2 shows a schematic diagram of a system for estimating advertisement delivery effect provided by an embodiment of the present specification, where the system includes an offline RTP module for executing steps 102 to 104, an offline playback module for executing steps 106 to 108, and a performance calibration module is further added in the system for calibrating the estimated delivery effect information, and obviously, the calibration method has two optional modes, one mode is to calibrate the estimated delivery effect information sequentially, and the other mode is to calibrate all the indicators once. For placement outcome prediction, the service only works when responding quickly to advertiser requests. Therefore, a multitask learning model is newly introduced into the delivery success estimation to calibrate the estimation deviation at one time, and meanwhile, the performance requirement and the accuracy requirement required by the estimation system are met. The model takes an MMOE model as a backbone, the model input consists of two parts, wherein an offline playback result is used as a basic estimation effect, and a plan setting is input into the model for learning a calibration mode. It is worth noting that the model maps the estimated basic result to the real delivery result in one time in an end-to-end mode, so that a plurality of calibration models do not need to be maintained frequently, estimated delivery effect information is calibrated through the calibration models, estimated results are more accurate, and the use experience of an advertiser is better.

In specific implementation, the offline RTP module acquires user characteristics and advertiser characteristics, scores recalled initial estimated flow through the offline RTP model, and sends the initial estimated flow carrying personalized scores to the offline playback module, the offline playback module determines target estimated flow based on the received initial estimated flow and the advertisement delivery plan, counts playback results (estimated delivery effect information) corresponding to the advertisement delivery plan, and inputs the playback results to the success calibration model for calibration, so as to obtain more accurate estimated delivery effect information.

Specifically, a technical architecture diagram of the offline RTP module can be seen in fig. 3, and fig. 3 illustrates a technical architecture diagram of the offline RTP model provided in an embodiment of the present specification; the Blink UDF module is used for executing tasks of determining initial estimated flow and scoring for each audience user, the Blink UDF module obtains user characteristics and advertiser characteristics from the characteristic center, then the user characteristics and the advertiser characteristics are input into the model service, and the model center in the model service calls a corresponding model to execute online computing service, so that the initial estimated flow is determined and scoring is conducted for each audience user.

The advertisement delivery effect estimation method provided by the specification comprises the steps of obtaining historical flow and advertisement delivery information, wherein the advertisement delivery information comprises bidding information; determining initial estimated flow according to the advertisement putting information and the historical flow; determining target estimated flow in the initial estimated flow according to the bidding information; and counting estimated delivery effect information of the advertisement delivery information based on the target estimated flow. By providing real-time estimated delivery effect information for the advertiser according to the historical flow and the advertisement delivery information, insight is provided for the advertiser, the advertiser can make an optimization decision aiming at the advertisement delivery information in advance, and trial and error cost is reduced.

The following will further explain the advertisement delivery effect estimation method by taking the application of the advertisement delivery effect estimation method provided in this specification to delivering commercial advertisements as an example with reference to fig. 4. Fig. 4 shows a flowchart of a processing procedure of an advertisement impression estimation method according to an embodiment of the present specification, and specific steps include step 402 to step 418.

Step 402: and obtaining historical flow and advertisement putting information to be put at this time.

The historical flow is the historical flow of the platform to be launched by the current advertisement acquired by the advertisement platform, and the advertisement launching information is a launching plan of the advertisement aiming at the commodity.

Step 404: and determining an audience user set according to the user label information in the advertisement putting information.

The user label information is attribute information expected to face the audience and preset in the advertisement putting information, and the audience user set is a user set which is screened from historical flow according to the user label information in the advertisement putting information and accords with the user label information.

Step 406: and scoring each audience user in the audience user set according to the commodity label information in the advertisement putting information to obtain the scoring information of each audience user.

The commodity label information is attribute information of the commodity of the advertisement, each audience user can be scored according to the commodity label information, and the scoring information comprises the click rate and the purchase rate of the commodity by the user.

Step 408: and determining initial estimated flow according to the flow information and the grading information of each audience user.

The initial estimated flow rate is all the estimated flow rates which can be obtained under the condition of not considering the flow rate bid.

Step 410: and determining the bidding type to be the CPC bidding type according to the bidding information in the advertisement delivery information.

The bidding type can be understood as a traffic bidding type and a budget set in the advertisement putting plan.

Step 412: and determining the target estimated flow in the initial estimated flow according to the unit price information in the bidding information.

The unit price information is the highest bid set in the advertisement putting plan, estimated traffic of which the historical unit price information is less than or equal to the unit price information is selected from the initial estimated traffic according to the unit price information, and all the selected estimated traffic is combined to generate target estimated traffic. The target estimated flow is all the estimated flows which can be obtained under the condition of considering the flow bid.

Step 414: and counting estimated delivery effect information of the advertisement delivery information based on the target estimated flow.

The estimated advertisement delivery effect information comprises a total display value estimated by the advertisement delivery information, a total click rate and a total cost.

Step 416: and calculating calibration information corresponding to the estimated putting effect information according to the estimated putting effect information and the advertisement putting information.

Step 418: and calibrating the estimated putting effect information according to the calibration information, and obtaining and feeding back the calibrated estimated putting effect information.

The specification provides an advertisement delivery effect estimation method applied to delivery of commodity advertisements, which comprises the steps of obtaining historical flow and advertisement delivery information to be delivered at this time, determining audience user sets according to user label information in the advertisement delivery information, scoring each audience user in the audience user sets according to commodity label information in the advertisement delivery information to obtain scoring information of each audience user, determining initial estimated flow according to the flow information and the scoring information of each audience user, determining a bidding type to be a CPC bidding type according to bidding information in the advertisement delivery information, determining target flow in the initial estimated flow according to unit price information in the bidding information, counting estimated delivery effect information of the advertisement delivery information based on the target estimated flow, calculating calibration information corresponding to the estimated delivery effect information according to the estimated delivery effect information and the advertisement delivery information, and calibrating the estimated putting effect information according to the calibration information, and obtaining and feeding back the calibrated estimated putting effect information. The method comprises the steps of obtaining target estimated flow carrying score information according to historical flow and advertisement delivery information, counting estimated delivery effect information of the advertisement delivery information according to the target estimated flow carrying the score information, calibrating the estimated delivery effect information, obtaining more accurate estimated delivery effect information and feeding the more accurate estimated delivery effect information back to an advertiser, providing advertisement delivery effect reference for the advertiser when the advertiser creates a new advertisement and modifies plan setting, based on the estimated delivery effect reference, enabling the advertiser to avoid blindly optimizing attempts, effectively shortening decision-making period and enabling the advertisement effect to reach the expected range of the advertiser as soon as possible.

Corresponding to the above method embodiment, the present specification further provides an advertisement delivery effect estimation apparatus embodiment, and fig. 5 shows a schematic structural diagram of an advertisement delivery effect estimation apparatus provided in an embodiment of the present specification. As shown in fig. 5, the apparatus includes:

an obtaining module 502 configured to obtain historical traffic and advertisement delivery information, wherein the advertisement delivery information includes bid information;

a first determining module 504 configured to determine an initial pre-estimated traffic according to the advertisement delivery information and the historical traffic;

a second determining module 506, configured to determine a target predicted traffic in the initial predicted traffic according to the bidding information;

a statistic module 508 configured to count estimated delivery effect information of the advertisement delivery information based on the target estimated traffic.

Optionally, the first determining module 504 is further configured to:

determining an audience user set according to the user tag information;

Optionally, the first determining module 504 is further configured to:

acquiring a historical user set corresponding to the historical traffic;

and determining an audience user set according to the screening result.

Optionally, the first determining module 504 is further configured to:

Optionally, the second determining module 506 is further configured to:

and determining the target estimated flow according to the estimated flow corresponding to at least one round of comparison results.

Optionally, the second determining module 506 is further configured to: the unit price bidding condition includes that the historical unit price information is less than or equal to the unit price information.

Optionally, the second determining module 506 is further configured to:

Optionally, the second determining module 506 is further configured to: the total bid bidding condition includes that historical total bid information is less than or equal to the total bid information.

Optionally, the statistics module 508 is further configured to: the target estimated traffic comprises score information of audience users corresponding to each estimated traffic;

Optionally, the apparatus further comprises a calibration module configured to:

The advertisement delivery effect estimation device provided by the specification comprises an acquisition module, a storage module and a display module, wherein the acquisition module is configured to acquire historical flow and advertisement delivery information, and the advertisement delivery information comprises bidding information; a first determination module configured to determine an initial estimated traffic according to the advertisement delivery information and the historical traffic; a second determination module configured to determine a target pre-estimated flow rate from the initial pre-estimated flow rates according to the bidding information; and the statistic module is configured to count estimated delivery effect information of the advertisement delivery information based on the target estimated flow. The method comprises the steps of obtaining target estimated flow carrying score information according to historical flow and advertisement delivery information, counting estimated delivery effect information of the advertisement delivery information according to the target estimated flow carrying the score information, calibrating the estimated delivery effect information, obtaining more accurate estimated delivery effect information and feeding the more accurate estimated delivery effect information back to an advertiser, providing advertisement delivery effect reference for the advertiser when the advertiser creates a new advertisement and modifies plan setting, based on the estimated delivery effect reference, enabling the advertiser to avoid blindly optimizing attempts, effectively shortening decision-making period and enabling the advertisement effect to reach the expected range of the advertiser as soon as possible.

The foregoing is a schematic scheme of an advertisement delivery effect estimation apparatus according to this embodiment. It should be noted that the technical scheme of the advertisement delivery effect estimation device and the technical scheme of the advertisement delivery effect estimation method belong to the same concept, and details of the technical scheme of the advertisement delivery effect estimation device, which are not described in detail, can be referred to the description of the technical scheme of the advertisement delivery effect estimation method.

Fig. 6 illustrates a block diagram of a computing device 600 provided according to an embodiment of the present description. The components of the computing device 600 include, but are not limited to, a memory 610 and a processor 620. The processor 620 is coupled to the memory 610 via a bus 630 and a database 650 is used to store data.

Computing device 600 also includes access device 640, access device 640 enabling computing device 600 to communicate via one or more networks 660. Examples of such networks include the Public Switched Telephone Network (PSTN), a Local Area Network (LAN), a Wide Area Network (WAN), a Personal Area Network (PAN), or a combination of communication networks such as the internet. Access device 640 may include one or more of any type of network interface (e.g., a Network Interface Card (NIC)) whether wired or wireless, such as an IEEE802.11 Wireless Local Area Network (WLAN) wireless interface, a worldwide interoperability for microwave access (Wi-MAX) interface, an ethernet interface, a Universal Serial Bus (USB) interface, a cellular network interface, a bluetooth interface, a Near Field Communication (NFC) interface, and so forth.

In one embodiment of the present description, the above-described components of computing device 600, as well as other components not shown in FIG. 6, may also be connected to each other, such as by a bus. It should be understood that the block diagram of the computing device architecture shown in FIG. 6 is for purposes of example only and is not limiting as to the scope of the present description. Those skilled in the art may add or replace other components as desired.

Computing device 600 may be any type of stationary or mobile computing device, including a mobile computer or mobile computing device (e.g., tablet, personal digital assistant, laptop, notebook, netbook, etc.), mobile phone (e.g., smartphone), wearable computing device (e.g., smartwatch, smartglasses, etc.), or other type of mobile device, or a stationary computing device such as a desktop computer or PC. Computing device 600 may also be a mobile or stationary server.

Wherein, the processor 620 implements the steps of the advertisement delivery effect estimation method when executing the computer instructions.

The above is an illustrative scheme of a computing device of the present embodiment. It should be noted that the technical solution of the computing device and the technical solution of the advertisement effectiveness estimating method belong to the same concept, and details that are not described in detail in the technical solution of the computing device can all be referred to in the description of the technical solution of the advertisement effectiveness estimating method.

An embodiment of the present specification further provides a computer readable storage medium, which stores computer instructions, and the computer instructions, when executed by a processor, implement the steps of the advertisement delivery effect estimation method as described above.

The above is an illustrative scheme of a computer-readable storage medium of the present embodiment. It should be noted that the technical solution of the storage medium and the technical solution of the advertisement effectiveness estimating method belong to the same concept, and details that are not described in detail in the technical solution of the storage medium can be referred to the description of the technical solution of the advertisement effectiveness estimating method.

An embodiment of the present specification further provides a computer program, wherein when the computer program is executed in a computer, the computer is caused to execute the steps of the advertisement delivery effect estimation method.

The above is a schematic scheme of a computer program of the present embodiment. It should be noted that the technical solution of the computer program and the technical solution of the advertisement effectiveness estimating method belong to the same concept, and details that are not described in detail in the technical solution of the computer program can be referred to the description of the technical solution of the advertisement effectiveness estimating method.

The foregoing description has been directed to specific embodiments of this disclosure. Other embodiments are within the scope of the following claims. In some cases, the actions or steps recited in the claims may be performed in a different order than in the embodiments and still achieve desirable results. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing may also be possible or may be advantageous.

The computer instructions comprise computer program code which may be in the form of source code, object code, an executable file or some intermediate form, or the like. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, usb disk, removable hard disk, magnetic disk, optical disk, computer Memory, Read-Only Memory (ROM), Random Access Memory (RAM), electrical carrier wave signals, telecommunications signals, software distribution medium, etc. It should be noted that the computer readable medium may contain content that is subject to appropriate increase or decrease as required by legislation and patent practice in jurisdictions, for example, in some jurisdictions, computer readable media does not include electrical carrier signals and telecommunications signals as is required by legislation and patent practice.

It should be noted that, for the sake of simplicity, the foregoing method embodiments are described as a series of acts, but those skilled in the art should understand that the present embodiment is not limited by the described acts, because some steps may be performed in other sequences or simultaneously according to the present embodiment. Further, those skilled in the art should also appreciate that the embodiments described in this specification are preferred embodiments and that acts and modules referred to are not necessarily required for an embodiment of the specification.

In the above embodiments, the descriptions of the respective embodiments have respective emphasis, and for parts that are not described in detail in a certain embodiment, reference may be made to related descriptions of other embodiments.

The preferred embodiments of the present specification disclosed above are intended only to aid in the description of the specification. Alternative embodiments are not exhaustive and do not limit the invention to the precise embodiments described. Obviously, many modifications and variations are possible in light of the above teaching. The embodiments were chosen and described in order to best explain the principles of the embodiments and the practical application, to thereby enable others skilled in the art to best understand and utilize the embodiments. The specification is limited only by the claims and their full scope and equivalents.

Claims

1. An advertisement putting effect estimation method comprises the following steps:

2. The method of claim 1, the advertisement placement information further comprising user tag information;

determining an audience user set according to the user tag information;

3. The method of claim 2, determining an audience user set from the user tag information, comprising:

acquiring a historical user set corresponding to the historical traffic;

and determining an audience user set according to the screening result.

4. The method of claim 2, the advertising information further comprising merchandise tag information;

5. The method of claim 1, determining a target predicted traffic among the initial predicted traffic based on bid information in the advertisement placement information, comprising:

6. The method of claim 5, determining a target predictive traffic volume in the initial predictive traffic volume based on unit price information in the bidding information, comprising:

7. The method of claim 6, wherein the unit price bidding condition comprises historical unit price information being less than or equal to the unit price information.

8. The method of claim 5, determining a target predictive traffic volume in the initial predictive traffic volume based on total price information in the bidding information, comprising:

9. The method of claim 8, wherein the total bid condition comprises historical total bid information being less than or equal to the total bid information.

10. The method of claim 1, wherein the target projected traffic includes rating information of audience users corresponding to each projected traffic;

11. The method of claim 1, further comprising:

12. An advertisement placement effect prediction device comprising:

13. A computing device comprising a memory, a processor, and computer instructions stored on the memory and executable on the processor, the processor implementing the steps of the method of any one of claims 1-11 when executing the computer instructions.

14. A computer-readable storage medium storing computer-executable instructions which, when executed by a processor, implement the steps of the method of any one of claims 1-11.

15. A computer program which, when executed on a computer, causes the computer to perform the steps of the advertisement impression prediction method according to any one of claims 1 to 11.