CN112446724A - Bidding method, device and equipment based on effect evaluation and readable storage medium - Google Patents
Bidding method, device and equipment based on effect evaluation and readable storage medium Download PDFInfo
- Publication number
- CN112446724A CN112446724A CN201910823275.8A CN201910823275A CN112446724A CN 112446724 A CN112446724 A CN 112446724A CN 201910823275 A CN201910823275 A CN 201910823275A CN 112446724 A CN112446724 A CN 112446724A
- Authority
- CN
- China
- Prior art keywords
- users
- group
- advertisement
- bid
- candidate
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0241—Advertisements
- G06Q30/0273—Determination of fees for advertising
- G06Q30/0275—Auctions
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Accounting & Taxation (AREA)
- Development Economics (AREA)
- Strategic Management (AREA)
- Finance (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Economics (AREA)
- Marketing (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a bidding method, a bidding device, bidding equipment and a readable storage medium based on effect evaluation. The method comprises the following steps: randomly simulating bids of the experimental advertisement to randomly divide the first group of users and the second group of users; acquiring behavior data of a first group of users and a second group of users; determining prior distribution functions of the first group of users and the second group of users with respect to the estimation parameters based on the acquired behavior data of the first group of users and the second group of users; comparing the prior distribution functions of the first group of users and the second group of users about the estimation parameters to determine the effect interval of the experimental advertisement; determining a first candidate bid according to the effect interval and the delivery cost of the experimental advertisement provided by the advertiser; determining whether a first candidate bid can meet a maximum profit transformation target in a preset frequency range of requests; and determining the first candidate bid as the optimal bid when the first candidate bid can meet the maximum revenue conversion target in the requests of the time threshold.
Description
Technical Field
The invention relates to an intelligent advertisement bidding technology, in particular to a bidding method, a bidding device, bidding equipment and a readable storage medium based on effect evaluation.
Background
Currently, in an advertisement program trading platform, intelligent bidding and advertisement effect evaluation experiments are independently performed, and a scheme integrating the intelligent bidding and advertisement effect evaluation experiments is not provided, so that an optimal bidding is given based on advertisement effect evaluation.
In the intelligent bidding scheme, the advertisement demander platform optimizes the advertisement click volume through an algorithm to perform intelligent bidding for the advertiser instead. The evaluation method cannot comprehensively and accurately evaluate the real effect of the advertisement because the standard is too single.
In an experiment for evaluating the advertisement effect, it is generally required that users are randomly divided into a test group and a control group in advance of the experiment, and the advertisement effect is evaluated by comparing click rates of the two groups of users. However, in the programmed transaction platform, both the advertiser and the advertisement demander platform cannot accurately know the identity of the potential user accessing the advertisement delivery platform, so that the potential access users cannot be randomly grouped accurately
Furthermore, in this experiment, the control group typically needs to be experimented with a public service advertisement instead of a commercial advertisement. Resulting in wasted advertiser budget. And because the public service advertisement and the commercial advertisement have different influences on the user behavior, it is difficult to ensure that the advertisement effect of the experimental evaluation has no deviation.
The above information disclosed in this background section is only for enhancement of understanding of the background of the invention and therefore it may contain information that does not constitute prior art that is already known to a person of ordinary skill in the art.
Disclosure of Invention
In view of the above, the present invention provides a bidding method, apparatus, device and readable storage medium based on effect evaluation.
Additional features and advantages of the invention will be set forth in the detailed description which follows, or may be learned by practice of the invention.
According to an aspect of the present invention, there is provided a bidding method based on effect evaluation, including: randomly simulating the bidding of the experimental advertisement within a preset first time length so as to randomly divide the potential users accessing the advertisement putting platform into a first group of users and a second group of users; respectively acquiring the behavior data of the first group of users and the second group of users within a preset second time length; respectively determining prior distribution functions of the first group of users and the second group of users with respect to an estimated parameter based on the acquired behavior data of the first group of users and the second group of users; comparing the prior distribution functions of the first group of users and the second group of users with respect to the estimated parameters to determine an effect interval of the experimental advertisement; determining a plurality of first candidate bids according to the effect interval and the putting cost of the experimental advertisement provided by the advertiser; determining whether one first candidate bid in the plurality of first candidate bids can meet the maximum profit transformation target in a preset number of requests; and determining the first candidate bid as an optimal bid when there are requests for which the first candidate bid can meet the maximum revenue conversion objective in each of the number of threshold requests; wherein the first group of users are users who can see the experimental advertisement in the advertisement delivery platform, and the second group of users are users who can not see the experimental advertisement in the advertisement delivery platform; and the maximum profit conversion target is determined according to the behavior data of the first group of users and the second group of users and the putting cost of the experimental advertisement.
According to an embodiment of the invention, the method further comprises: when there is no first candidate bid that can satisfy the maximum revenue conversion objective in each of the number of threshold requests, performing the following: acquiring real-time feedback data of the first group of users and the second group of users within a preset third time length from the advertisement putting platform; determining a posterior distribution function related to the estimation parameter according to the real-time feedback data; determining the probability of each first candidate bid becoming the optimal bid according to the posterior distribution function and the revenue conversion function based on the first candidate bids; determining whether the probability of the presence of a first candidate bid reaches a preset probability threshold; and determining the first candidate bid as an optimal bid when the probability that the first candidate bid exists reaches the probability threshold.
According to an embodiment of the invention, the method further comprises: when the probability that the first candidate bid does not exist reaches the probability threshold, performing the following: updating the plurality of first candidate bids into a plurality of second candidate bids according to the posterior distribution function; according to the plurality of second candidate bids, bidding on the advertisement putting platform; acquiring real-time feedback data of the first group of users and the second group of users within a preset fourth time length from the advertisement putting platform; updating a posterior distribution function related to the estimation parameter according to the real-time feedback data; determining the probability of each second candidate bid becoming the optimal bid according to the updated posterior distribution function and the income conversion function based on the second candidate bids; determining whether said probability of the presence of a second candidate bid meets said probability threshold; and determining the second candidate bid as an optimal bid when the probability that the second candidate bid exists reaches the probability threshold.
According to an embodiment of the present invention, the estimating parameters respectively include: the mean and variance of the behavioral data of the first group of users, and the mean and variance of the behavioral data of the second group of users.
According to an embodiment of the invention, the a priori distribution function with respect to the estimation parameter comprises: and a gamma distribution function and a normal distribution function estimated based on a Bayesian method.
According to an embodiment of the invention, the revenue conversion function is determined based on the maximum revenue conversion objective.
According to an embodiment of the invention, the real-time feedback data comprises at least one of the following data: exposure data, click service data, and order data.
According to another aspect of the present invention, there is provided a bidding apparatus based on effectiveness evaluation, including: the random bidding module is used for randomly simulating bidding of the experimental advertisement within a preset first time length so as to randomly divide potential users accessing the advertisement putting platform into a first group of users and a second group of users; the data acquisition module is used for respectively acquiring the behavior data of the first group of users and the second group of users within a preset second time length; a distribution determination module, configured to determine, based on the obtained behavior data of the first group of users and the second group of users, prior distribution functions of the first group of users and the second group of users with respect to an estimation parameter, respectively; a distribution comparison module, configured to compare prior distribution functions of the first group of users and the second group of users with respect to the estimation parameter, so as to determine an effect interval of the experimental advertisement; the bid determination module is used for determining a plurality of first candidate bids according to the effect interval and the putting cost of the experimental advertisement provided by the advertiser; the effect evaluation module is used for determining whether one first candidate bid in the plurality of first candidate bids can meet the goal of maximum profit conversion in a preset frequency range of requests; and an optimal bid module for determining the first candidate bid as an optimal bid when there is a first candidate bid that can satisfy the maximum revenue conversion objective in the number of threshold requests; wherein the first group of users are users who can see the experimental advertisement in the advertisement delivery platform, and the second group of users are users who can not see the experimental advertisement in the advertisement delivery platform; and the maximum profit conversion target is determined according to the behavior data of the first group of users and the second group of users and the putting cost of the experimental advertisement.
According to still another aspect of the present invention, there is provided a computer apparatus comprising: a memory, a processor and executable instructions stored in the memory and executable in the processor, the processor implementing any of the methods described above when executing the executable instructions.
According to yet another aspect of the present invention, there is provided a computer-readable storage medium having stored thereon computer-executable instructions which, when executed by a processor, implement any of the methods described above.
According to the bidding method based on effect evaluation provided by the embodiment of the invention, a scheme integrating intelligent bidding and advertisement effect evaluation experiments is provided, and the optimal bidding can be given based on advertisement effect evaluation, namely the advertisement effect evaluation is realized by adjusting the optimal bidding. In the method, the advertisement effect is used as an evaluation basis of intelligent bidding, and compared with a scheme of carrying out intelligent bidding based on click rate in the related art, the method can evaluate the real effect of the advertisement more comprehensively and accurately; in addition, the method avoids the problem that the potential users need to be grouped manually in advance in the related technology through random bidding, and provides a more accurate user grouping mode. In addition, the control group does not need to use public service advertisements for evaluation, so that not only is the waste of budget of an advertiser avoided, but also the deviation of the advertisement effect of experimental evaluation is ensured.
It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the invention, as claimed.
Drawings
The above and other objects, features and advantages of the present invention will become more apparent by describing in detail exemplary embodiments thereof with reference to the attached drawings.
FIG. 1 is a block diagram illustrating a system for programmatic trading of advertisements according to an example.
FIG. 2 is a flow diagram illustrating advertisement processing in the real-time bid server 13 according to an example.
Fig. 3 is a schematic diagram illustrating the partitioning of experimental stages according to an example.
FIG. 4 is a flow diagram illustrating a method of bidding based on effectiveness evaluation according to an exemplary embodiment.
FIG. 5 is a flow diagram illustrating another effectiveness evaluation-based bidding method according to an exemplary embodiment.
Fig. 6 is a flow diagram illustrating yet another effectiveness evaluation-based bidding method according to an exemplary embodiment.
FIG. 7 is a block diagram illustrating an effectiveness evaluation-based bidding appliance, according to an exemplary embodiment.
FIG. 8 is a block diagram illustrating another effectiveness evaluation-based bidding appliance, according to an exemplary embodiment.
FIG. 9 is a block diagram illustrating a computer system in accordance with an exemplary embodiment.
Detailed Description
Example embodiments will now be described more fully with reference to the accompanying drawings. Example embodiments may, however, be embodied in many different forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the concept of example embodiments to those skilled in the art. The drawings are merely schematic illustrations of the invention and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and thus their repetitive description will be omitted.
Furthermore, the described features, structures, or characteristics may be combined in any suitable manner in one or more embodiments. In the following description, numerous specific details are provided to provide a thorough understanding of embodiments of the invention. One skilled in the relevant art will recognize, however, that the invention may be practiced without one or more of the specific details, or with other methods, components, devices, steps, and so forth. In other instances, well-known structures, methods, devices, implementations, or operations are not shown or described in detail to avoid obscuring aspects of the invention.
Furthermore, in the description of the present invention, the terms "first" and "second" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implying any number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature.
FIG. 1 is a block diagram illustrating a system for programmatic trading of advertisements according to an example.
Referring to fig. 1, the system 1 comprises: an advertisement delivery platform 11, a programmed trading platform 12 and a real-time bidding server 13.
Wherein, the advertisement putting platform 11 may further include: a front-end client and a back-end server. The client of the front end may be, for example, a browser based on a PC, or may also be an application (including an applet based on instant messaging software such as a WeChat) based on an intelligent terminal device (such as a smart phone, a PAD, and the like).
The programmatic trading platform 12 communicates with the real-time bidding server 13 and sends information of the advertisement to be delivered or the experimental advertisement to the real-time bidding server 13. The real-time bidding server 13 performs a conventional bidding on the advertisement to be delivered, or performs a bidding method based on effect evaluation described below on the experimental advertisement, so as to provide an optimal bidding for the experimental advertisement based on the advertisement effect.
The real-time bidding server 13 may also communicate with the advertising platform 11 to obtain real-time user feedback data for conventional bidding as described above or for optimal bidding on experimental advertisements.
The real-time bidding server 13 may be a single server or a distributed cluster server, for example, but the present invention is not limited thereto.
FIG. 2 is a flow diagram illustrating advertisement processing in the real-time bid server 13 according to an example.
Referring to fig. 2, when the real-time bidding server 13 receives the advertisement requested by the programmatic trading platform, it determines whether the advertisement is the experimental advertisement (step S21). If not a trial advertisement, the conventional bidding logic described above is executed (step S22). If it is a trial advertisement, it is determined that the trial advertisement is currently in the stage (step S23). It is determined whether the experimental advertisement is in the first stage of the experiment (step S24). If it is in the first stage of the experiment, the experimental advertisement is randomly bid (step S25). If not, it is determined whether the second stage of the experiment is performed (step S26). If it is not in the second stage of the experiment, the conventional bidding logic described above is executed (step S22). If it is in the second stage of the experiment, it is determined whether the experiment finds the optimal bid (step S27). If an optimal bid is found, bidding is stopped and placement of the experimental ad is terminated (step S28). If no optimal bid is found, the price in the posterior data is used for bidding (step S29), and the flow is returned to the beginning.
Fig. 3 is a schematic diagram illustrating the partitioning of experimental stages according to an example. As shown in fig. 3, in the first stage of experiment, a random bidding manner is adopted to model the collected data of different groups of users; and in the second stage of the experiment, updating the model according to the collected user feedback data, and bidding according to the updated model until the optimal bidding meeting the requirements is found.
Specific bidding methods based on effectiveness evaluation will be further described below.
FIG. 4 is a flow diagram illustrating a method of bidding based on effectiveness evaluation according to an exemplary embodiment. This method may be performed, for example, in the real-time bid server 13 described above.
Referring to fig. 3, the bidding method 10 based on the effectiveness evaluation includes:
in step S102, randomly simulating bids of experimental advertisements within a preset first time period to randomly divide potential users accessing the advertisement delivery platform into a first group of users and a second group of users.
The first group of users are users who can see the experimental advertisement in the advertisement putting platform, and the second group of users are users who can not see the experimental advertisement in the advertisement putting platform.
Because the experimental advertisement is bid randomly, the experimental advertisement may be auctioned to release the experimental advertisement, or the experimental advertisement may not be auctioned to release the experimental advertisement. For the case of successfully publishing the experimental advertisement, the obtained users (users who can see the experimental advertisement) may be divided into a first group (such as the aforementioned test group); in the case where the experimental advertisement is not successfully distributed, the acquired users (users who cannot see the experimental advertisement) are divided into a second group (such as the aforementioned control group).
By randomly bidding on the experimental advertisements, the purpose of randomly classifying users into different groups can be achieved. Compared with the experimental scheme for evaluating the advertising effect in the related art, the potential users do not need to be manually grouped in advance, and the division mode is more accurate.
The first preset time period may be, for example, 12 hours as shown in fig. 3, but the invention is not limited thereto, and the length thereof may be set according to actual requirements in practical application.
In step S104, behavior data of the first group of users and behavior data of the second group of users are respectively obtained within a preset second duration.
The second period of time may be, for example, 24 hours in the period B in the first stage shown in fig. 3. And respectively acquiring the behavior data of the first group of users and the second group of users in the time length.
The behavior data may be, for example, purchase data of different groups of users, such as order information of purchased goods, and the like.
The behavioural data of the first group of users and the behavioural data of the second group of users may for example be denoted as Y (1), Y (0), respectively.
In step S106, based on the acquired behavior data of the first group of users and the second group of users, a prior distribution function of the first group of users and the second group of users with respect to the estimated parameter is determined, respectively.
First, behavioral data for a first group of users and a second group of users are modeled, respectively. Assume that the behavior data Y (1) of the first group of users and the behavior data Y (0) of the second group of users follow the following distribution:
where i is used to represent different user data. B iscp，iUser behavior data representing the ith user at the highest bid.
The estimated parameters θ may include, for example: mean and variance of the behavioral data of the first group of users, mean (δ) and variance (σ) of the behavioral data of the second group of users2). That is, the parameters are estimated
Secondly, estimating by using a Bayesian method, and for k ∈ {1, 0, CP }, a prior distribution function p (theta) related to an estimation parameter theta is as follows:
wherein, formula (1) is a gamma distribution function, and formula (2) is a normal distribution function.
Parameter α in a prior distribution functionk，βk，
May be calculated using the mean and variance of the behavioral data of the first group of users and the behavioral data of the second group of users.
In step S108, the prior distribution functions of the first group of users and the second group of users with respect to the estimated parameters are compared to determine the effectiveness interval of the experimental advertisement.
The effectiveness of the experimental advertisement can be measured according to the formula Y (1) -Y (0).
In step S110, a plurality of first candidate bids are determined according to the effect interval of the experimental advertisement and the placement cost of the experimental advertisement provided by the advertiser.
For example, a numerical interval can be determined according to the advertisement effect interval and the thousands of showing costs provided by the advertiser, the interval is divided into a plurality of equal parts on average, and the boundary of each equal part is a candidate bid.
Several first candidate bids, for example, can be represented as [ b ]1，b2，...，bn]Where n is the number of first candidate bids.
The placement Cost of the experimental advertisement may be thousands of presentation costs (Cost Per mill, CPM), for example.
In step S112, it is determined whether there is one of the first candidate bids among the first candidate bids that can satisfy the maximum profit-conversion goal in a preset number of requests.
The maximum revenue conversion goal may be determined, for example, based on behavioral data of the first and second groups of users and placement costs of the experimental advertisement. For example, the maximum revenue transformation goal can be obtained by calculating the sum and difference of the behavior data of all the first group users and all the second group users, and then subtracting the placement cost of the experimental advertisement.
The predetermined number of times may be, for example, 95%, that is, it is determined whether there is a first candidate bid that can satisfy the maximum profit-conversion target in 95% of the requests, but the present invention is not limited thereto. The request may be, for example, a bid request for the experimental advertisement.
In step S114, when there is a first candidate bid that can satisfy the maximum benefit conversion target in each of the requests of the number threshold, the first candidate bid is determined to be the optimal bid.
If the first candidate bid exists, the first candidate bid is determined to be the optimal bid, and the method is ended. Otherwise, the second phase of the experiment is entered.
According to the bidding method based on effect evaluation provided by the embodiment of the invention, a scheme integrating intelligent bidding and advertisement effect evaluation experiments is provided, and the optimal bidding can be given based on advertisement effect evaluation, namely the advertisement effect evaluation is realized by adjusting the optimal bidding. In the method, the advertisement effect is used as an evaluation basis of intelligent bidding, and compared with a scheme of carrying out intelligent bidding based on click rate in the related art, the method can evaluate the real effect of the advertisement more comprehensively and accurately; in addition, the method avoids the problem that the potential users need to be grouped manually in advance in the related technology through random bidding, and provides a more accurate user grouping mode. In addition, the control group does not need to use public service advertisements for evaluation, so that not only is the waste of budget of an advertiser avoided, but also the deviation of the advertisement effect of experimental evaluation is ensured.
It should be clearly understood that the present disclosure describes how to make and use particular examples, but the principles of the present disclosure are not limited to any details of these examples. Rather, these principles can be applied to many other embodiments based on the teachings of the present disclosure.
FIG. 5 is a flow diagram illustrating another effectiveness evaluation-based bidding method according to an exemplary embodiment. This method may be performed, for example, in the real-time bid server 13 described above.
Referring to FIG. 5, the effectiveness evaluation-based bidding method 20 shown in FIG. 5 further provides an implementation when the first candidate bid is not present, as compared to the effectiveness evaluation-based bidding method 10 shown in FIG. 4, the method 20 as may be implemented in the experimental second phase shown in FIG. 3. Compared to method 10, method 20 further comprises:
in step S202, real-time feedback data of the first group of users and the second group of users within a preset third duration is obtained from the advertisement delivery platform.
The third preset time period may be, for example, the time period of the second phase of the experiment shown in fig. 3, which is 2 hours, and the real-time feedback data of the first group of users and the second group of users of the second phase of the experiment are collected at the time period D shown in fig. 3.
The real-time feedback data may for example comprise at least one of the following data: and the exposure data, the click service data and the order data fed back by the user in real time.
In step S204, a posterior distribution function with respect to the estimated parameters is determined based on the real-time feedback data.
For example, the feedback data may be expressed as data (x), and the estimation parameter θ may be estimated by bayesian rules to obtain a more accurate posterior distribution p (θ | x).
It can be obtained by using bayesian formula:
where p (x | θ) is the likelihood function, p (θ) is given in method 10. Therefore, the posterior distribution p (θ | x) of θ can be obtained by calculation.
The mean value δ of the behavioral data of the first group of users is given below1For example, how to calculate the posterior distribution function of the estimated parameters is specified according to equation (5).
In the formula (5), S is a data matrix, Y is behavior data of a first group of users, N is a sample size, and δ and σ are parameters in the prior distribution function p (θ). The posterior distribution can be understood as being the weighted average of the prior distribution and the real-time feedback data.
The posterior distribution functions of other parameters can be calculated by the same method, and are not described in detail herein.
In step S206, a probability of each first candidate bid becoming the optimal bid is determined according to a posterior distribution function and an revenue conversion function based on the first candidate bids.
For example, sets of simulation data may be generated based on a posterior distribution function, with optimal bids in each set of simulation data determined in accordance with an revenue transformation objective function. In this way, the proportion of each candidate bid that becomes the optimal bid is calculated, which is the probability that the candidate bid becomes the optimal bid.
Wherein the revenue transformation function may be determined based on the maximum revenue transformation objective described above.
In step S208, it is determined whether the probability that there is a first candidate bid reaches a preset probability threshold.
The probability threshold may be, for example, 95%, but the invention is not limited thereto.
In step S210, the first candidate bid is determined to be the optimal bid when the probability that there is the first candidate bid reaches a probability threshold.
When the probability of finding a first candidate bid reaches the probability threshold, the first candidate bid is determined to be the optimal bid, and the method is ended.
According to the bidding method based on effect evaluation provided by the embodiment of the invention, a method for determining a posterior distribution function with more accurate estimated parameters in the second stage of an experiment by combining with real-time feedback data of a user and searching for the optimal bid according to the posterior distribution function is further provided.
Fig. 6 is a flow diagram illustrating yet another effectiveness evaluation-based bidding method according to an exemplary embodiment. This method may be performed, for example, in the real-time bid server 13 described above.
Referring to FIG. 6, the effectiveness evaluation-based bidding method 30 of FIG. 6 further provides an implementation when the probability of the absence of a first candidate bid reaches a probability threshold, as compared to the effectiveness evaluation-based bidding method 20 of FIG. 5, and the method 30 may still be implemented in the second phase of the experiment of FIG. 3. Thus, compared to method 20, method 30 further comprises:
in step S302, a number of first candidate bids are updated to a number of second candidate bids according to a posterior distribution function.
For example, after updating the posterior distribution function in conjunction with the user real-time feedback data, several sets of simulation data may be generated based on the updated posterior distribution function, and the probability that the first set of candidate bids will become the optimal bids may be calculated and updated according to the method in step S206 of the method 20 described above. And using the adjusted first set of candidate bids as a second set of candidate bids.
In step S304, bids are placed on the advertisement delivery platform according to the number of second candidate bids.
After determining a number of second candidate bids, bids are placed on the advertising platform according to the second candidate bids.
In step S306, real-time feedback data of the first group of users and the second group of users within a preset fourth duration is obtained from the advertisement delivery platform.
The fourth preset time period may be, for example, the time period E of the second stage of the experiment shown in fig. 3, which is 2 hours, and the real-time feedback data of the first group of users and the second group of users of the time period E is collected at the time period F shown in fig. 3.
The real-time feedback data may for example comprise at least one of the following data: and the exposure data, the click service data and the order data fed back by the user in real time.
In step S308, the posterior distribution function with respect to the estimated parameters is updated according to the real-time feedback data.
The calculation of the posterior distribution function is as described in method 20, and will not be described herein.
In step S310, the probability of each second candidate bid becoming the optimal bid is determined according to the updated posterior distribution function and the revenue conversion function based on the second candidate bids.
For example, as described above, sets of simulation data may be generated based on a posterior distribution function, with optimal bids in each set of simulation data determined in accordance with an revenue transformation objective function. In this way, the proportion of each candidate bid that becomes the optimal bid is calculated, which is the probability that the candidate bid becomes the optimal bid.
In step S312, it is determined whether the probability that there is a second candidate bid meets a probability threshold.
The probability threshold may be, for example, 95%, but the invention is not limited thereto.
In step S314, the second candidate bid is determined to be the optimal bid when the probability that there is the second candidate bid reaches a probability threshold.
And when the probability of finding a second candidate bid reaches the probability threshold, determining the second candidate bid as the optimal bid, and ending the method.
If the second candidate bid is not found, the method 30 can continue to be performed, updating the posterior probability distribution of the estimated parameters according to the user real-time feedback data, until an optimal bid meeting the above requirements is found.
Those skilled in the art will appreciate that all or part of the steps implementing the above embodiments are implemented as computer programs executed by a CPU. The computer program, when executed by the CPU, performs the functions defined by the method provided by the present invention. The program may be stored in a computer readable storage medium, which may be a read-only memory, a magnetic or optical disk, or the like.
Furthermore, it should be noted that the above-mentioned figures are only schematic illustrations of the processes involved in the method according to exemplary embodiments of the invention, and are not intended to be limiting. It will be readily understood that the processes shown in the above figures are not intended to indicate or limit the chronological order of the processes. In addition, it is also readily understood that these processes may be performed synchronously or asynchronously, e.g., in multiple modules.
The following are embodiments of the apparatus of the present invention that may be used to perform embodiments of the method of the present invention. For details which are not disclosed in the embodiments of the apparatus of the present invention, reference is made to the embodiments of the method of the present invention.
FIG. 7 is a block diagram illustrating an effectiveness evaluation-based bidding appliance, according to an exemplary embodiment.
Referring to fig. 7, the bidding apparatus 40 based on the effectiveness evaluation includes: a random bid module 402, a data acquisition module 404, a distribution determination module 406, a distribution comparison module 408, a bid determination module 410, an effect evaluation module 412, an optimal bid module 414.
The random bidding module 402 is configured to randomly simulate bidding of the experimental advertisement within a preset first time period, so as to randomly divide potential users accessing the advertisement delivery platform into a first group of users and a second group of users.
The data obtaining module 404 is configured to obtain behavior data of the first group of users and the second group of users within a preset second duration.
The distribution determination module 406 is configured to determine an a priori distribution function of the first group of users and the second group of users with respect to the estimated parameter, respectively, based on the obtained behavior data of the first group of users and the second group of users.
The distribution comparison module 408 is configured to compare the prior distribution functions of the first group of users and the second group of users with respect to the estimated parameters to determine the effectiveness interval of the experimental advertisement.
The bid determination module 410 is configured to determine a plurality of first candidate bids according to the effect interval and the delivery cost of the experimental advertisement provided by the advertiser.
The effectiveness evaluation module 412 is configured to determine whether, among a number of first candidate bids, there is one that satisfies the maximum revenue conversion goal within a preset range of times of the request.
The optimal bid module 414 is configured to determine the first candidate bid as the optimal bid when there is a first candidate bid that satisfies the maximum revenue conversion objective for each of the number of threshold requests.
The first group of users are users who can see the experimental advertisement in the advertisement putting platform, and the second group of users are users who can not see the experimental advertisement in the advertisement putting platform;
and determining the maximum profit transformation target according to the behavior data of the first group of users and the second group of users and the delivery cost of the experimental advertisement.
In some embodiments, estimating the parameters includes: the mean and variance of the behavioral data of the first group of users, and the mean and variance of the behavioral data of the second group of users.
In some embodiments, the prior distribution function comprises: and a gamma distribution function and a normal distribution function estimated based on a Bayesian method.
According to the bidding device based on effect evaluation provided by the embodiment of the invention, a scheme integrating intelligent bidding and advertisement effect evaluation experiments is provided, and the optimal bidding can be given based on advertisement effect evaluation, namely the advertisement effect evaluation can be realized by adjusting the optimal bidding. In the method, the advertisement effect is used as an evaluation basis of intelligent bidding, and compared with a scheme of carrying out intelligent bidding based on click rate in the related art, the method can evaluate the real effect of the advertisement more comprehensively and accurately; in addition, the method avoids the problem that the potential users need to be grouped manually in advance in the related technology through random bidding, and provides a more accurate user grouping mode. In addition, the control group does not need to use public service advertisements for evaluation, so that not only is the waste of budget of an advertiser avoided, but also the deviation of the advertisement effect of experimental evaluation is ensured.
FIG. 8 is a block diagram illustrating another effectiveness evaluation-based bidding appliance, according to an exemplary embodiment.
Referring to fig. 8, the effect evaluation-based bidding apparatus 40 shown in fig. 7 is different in that the effect evaluation-based bidding apparatus 50 shown in fig. 8 further includes, compared to the apparatus 40: a feedback acquisition module 502, a posterior distribution determination module 504, a probability determination module 506, a bid determination module 508, and a second optimal bid module 510.
The feedback obtaining module 502 is configured to obtain real-time feedback data of the first group of users and the second group of users within a preset third time period from the advertisement delivery platform when there is no first candidate bid that can meet the maximum profit transformation target in the request of the time threshold.
The posterior distribution determining module 504 is configured to determine a posterior distribution function with respect to the estimated parameter according to the real-time feedback data.
The probability determination module 506 is configured to determine a probability that each of the first candidate bids will become an optimal bid based on a posterior distribution function and an revenue conversion function based on the first candidate bids.
The bid selection module 508 is configured to determine whether a probability of a first candidate bid reaching a predetermined probability threshold.
The second optimal bidding module 510 is configured to determine the first candidate bid as the optimal bid when the probability that there is the first candidate bid reaches a probability threshold.
In some embodiments, the apparatus 50 may further comprise: a second bid determination module and a platform bid module. The second bid determination module is used for updating a plurality of first candidate bids into a plurality of second candidate bids according to a posterior distribution function when the probability that the first candidate bids do not exist reaches a probability threshold. And the platform bidding module is used for bidding on the advertisement putting platform according to the plurality of second candidate bids. The feedback obtaining module 502 is further configured to obtain real-time feedback data of the first group of users and the second group of users within a preset fourth time period from the advertisement delivery platform. The posterior distribution determining module 504 is further configured to update the posterior distribution function with respect to the estimated parameter according to the real-time feedback data. The probability determination module 506 is further configured to determine a probability that each second candidate bid will become the optimal bid based on the updated posterior distribution function and the revenue conversion function based on the second candidate bids. The bid selection module 508 is further operable to determine whether a probability that there is a second candidate bid meets a probability threshold. The second optimal bidding module 510 is also configured to determine the second candidate bid as the optimal bid when the probability that there is the second candidate bid reaches a probability threshold.
In some embodiments, the revenue conversion function is determined based on a maximum revenue conversion objective.
It is noted that the block diagrams shown in the above figures are functional entities and do not necessarily correspond to physically or logically separate entities. These functional entities may be implemented in the form of software, or in one or more hardware modules or integrated circuits, or in different networks and/or processor devices and/or microcontroller devices.
FIG. 9 is a block diagram illustrating a computer system in accordance with an exemplary embodiment. It should be noted that the computer system shown in fig. 9 is only an example, and should not bring any limitation to the functions and the scope of the application of the embodiments.
As shown in fig. 9, the computer system 800 includes a Central Processing Unit (CPU)801 that can perform various appropriate actions and processes in accordance with a program stored in a Read Only Memory (ROM)802 or a program loaded from a storage section 808 into a Random Access Memory (RAM) 803. In the RAM 803, various programs and data necessary for the operation of the system 800 are also stored. The CPU 801, ROM 802, and RAM 803 are connected to each other via a bus 804. An input/output (I/O) interface 805 is also connected to bus 804.
The following components are connected to the I/O interface 805: an input portion 806 including a keyboard, a mouse, and the like; an output section 807 including a signal such as a Cathode Ray Tube (CRT), a Liquid Crystal Display (LCD), and the like, and a speaker; a storage portion 808 including a hard disk and the like; and a communication section 809 including a network interface card such as a LAN card, a modem, or the like. The communication section 809 performs communication processing via a network such as the internet. A drive 810 is also connected to the I/O interface 805 as necessary. A removable medium 811 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, or the like is mounted on the drive 810 as necessary, so that a computer program read out therefrom is mounted on the storage section 808 as necessary.
In particular, according to an embodiment of the present disclosure, the processes described above with reference to the flowcharts may be implemented as computer software programs. For example, embodiments of the present disclosure include a computer program product comprising a computer program embodied on a computer readable medium, the computer program comprising program code for performing the method illustrated in the flow chart. In such an embodiment, the computer program can be downloaded and installed from a network through the communication section 809 and/or installed from the removable medium 811. The computer program executes the above-described functions defined in the system of the present application when executed by the Central Processing Unit (CPU) 801.
It should be noted that the computer readable medium shown in the present application may be a computer readable signal medium or a computer readable storage medium or any combination of the two. A computer readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the foregoing. More specific examples of the computer readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the present application, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. In this application, however, a computer readable signal medium may include a propagated data signal with computer readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated data signal may take many forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A computer readable signal medium may also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wire, fiber optic cable, RF, etc., or any suitable combination of the foregoing.
The flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present application. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams or flowchart illustration, and combinations of blocks in the block diagrams or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
The units described in the embodiments of the present application may be implemented by software or hardware. The described units may also be provided in a processor, and may be described as: a processor includes a transmitting unit, an obtaining unit, a determining unit, and a first processing unit. The names of these units do not in some cases constitute a limitation to the unit itself, and for example, the sending unit may also be described as a "unit sending a picture acquisition request to a connected server".
As another aspect, the present application also provides a computer-readable medium, which may be contained in the apparatus described in the above embodiments; or may be separate and not incorporated into the device. The computer readable medium carries one or more programs which, when executed by a device, cause the device to comprise:
randomly simulating the bidding of the experimental advertisement within a preset first time length so as to randomly divide the potential users accessing the advertisement putting platform into a first group of users and a second group of users;
respectively acquiring behavior data of a first group of users and a second group of users within a preset second time length;
respectively determining prior distribution functions of the first group of users and the second group of users about the estimation parameters based on the acquired behavior data of the first group of users and the second group of users;
comparing the prior distribution functions of the first group of users and the second group of users about the estimation parameters to determine the effect interval of the experimental advertisement;
determining a plurality of first candidate bids according to the effect interval and the putting cost of the experimental advertisement provided by the advertiser;
determining whether one first candidate bid in a plurality of first candidate bids can meet the maximum profit transformation target in a preset frequency range of requests; and
determining the first candidate bid as the optimal bid when the first candidate bid can meet the maximum profit transformation target in the requests of the times threshold;
the first group of users are users who can see the experimental advertisement in the advertisement putting platform, and the second group of users are users who can not see the experimental advertisement in the advertisement putting platform;
and determining the maximum profit transformation target according to the behavior data of the first group of users and the second group of users and the delivery cost of the experimental advertisement.
Exemplary embodiments of the present invention are specifically illustrated and described above. It is to be understood that the invention is not limited to the precise construction, arrangements, or instrumentalities described herein; on the contrary, the invention is intended to cover various modifications and equivalent arrangements included within the spirit and scope of the appended claims.
Claims (10)
1. A bidding method based on effect evaluation, comprising:
randomly simulating the bidding of the experimental advertisement within a preset first time length so as to randomly divide the potential users accessing the advertisement putting platform into a first group of users and a second group of users;
respectively acquiring the behavior data of the first group of users and the second group of users within a preset second time length;
respectively determining prior distribution functions of the first group of users and the second group of users with respect to an estimated parameter based on the acquired behavior data of the first group of users and the second group of users;
comparing the prior distribution functions of the first group of users and the second group of users with respect to the estimated parameters to determine an effect interval of the experimental advertisement;
determining a plurality of first candidate bids according to the effect interval and the putting cost of the experimental advertisement provided by the advertiser;
determining whether one first candidate bid in the plurality of first candidate bids can meet the maximum profit transformation target in a preset number of requests; and
determining the first candidate bid as an optimal bid when there are requests for which the first candidate bid can meet the maximum revenue conversion objective in each of the number of threshold requests;
wherein the first group of users are users who can see the experimental advertisement in the advertisement delivery platform, and the second group of users are users who can not see the experimental advertisement in the advertisement delivery platform;
and the maximum profit conversion target is determined according to the behavior data of the first group of users and the second group of users and the putting cost of the experimental advertisement.
2. The method of claim 1, further comprising: when there is no first candidate bid that can satisfy the maximum revenue conversion objective in each of the number of threshold requests, performing the following:
acquiring real-time feedback data of the first group of users and the second group of users within a preset third time length from the advertisement putting platform;
determining a posterior distribution function related to the estimation parameter according to the real-time feedback data;
determining the probability of each first candidate bid becoming the optimal bid according to the posterior distribution function and the revenue conversion function based on the first candidate bids;
determining whether the probability of the presence of a first candidate bid reaches a preset probability threshold; and
determining that the first candidate bid is an optimal bid when the probability that the first candidate bid exists reaches the probability threshold.
3. The method of claim 2, further comprising: when the probability that the first candidate bid does not exist reaches the probability threshold, performing the following:
updating the plurality of first candidate bids into a plurality of second candidate bids according to the posterior distribution function;
according to the plurality of second candidate bids, bidding on the advertisement putting platform;
acquiring real-time feedback data of the first group of users and the second group of users within a preset fourth time length from the advertisement putting platform;
updating a posterior distribution function related to the estimation parameter according to the real-time feedback data;
determining the probability of each second candidate bid becoming the optimal bid according to the updated posterior distribution function and the income conversion function based on the second candidate bids;
determining whether said probability of the presence of a second candidate bid meets said probability threshold; and
determining the second candidate bid as an optimal bid when the probability that the second candidate bid exists reaches the probability threshold.
4. The method according to any of claims 1-3, wherein estimating the parameters comprises: the mean and variance of the behavioral data of the first group of users, and the mean and variance of the behavioral data of the second group of users.
5. The method of claim 4, wherein the a priori distribution function for the estimation parameter comprises: and a gamma distribution function and a normal distribution function estimated based on a Bayesian method.
6. The method of any of claims 1-3, wherein the revenue conversion function is determined based on the maximum revenue conversion objective.
7. A method according to any of claims 1-3, wherein the real-time feedback data comprises at least one of: exposure data, click service data, and order data.
8. A bidding device based on effectiveness evaluation, comprising:
the random bidding module is used for randomly simulating bidding of the experimental advertisement within a preset first time length so as to randomly divide potential users accessing the advertisement putting platform into a first group of users and a second group of users;
the data acquisition module is used for respectively acquiring the behavior data of the first group of users and the second group of users within a preset second time length;
a distribution determination module, configured to determine, based on the obtained behavior data of the first group of users and the second group of users, prior distribution functions of the first group of users and the second group of users with respect to an estimation parameter, respectively;
a distribution comparison module, configured to compare prior distribution functions of the first group of users and the second group of users with respect to the estimation parameter, so as to determine an effect interval of the experimental advertisement;
the bid determination module is used for determining a plurality of first candidate bids according to the effect interval and the putting cost of the experimental advertisement provided by the advertiser;
the effect evaluation module is used for determining whether one first candidate bid in the plurality of first candidate bids can meet the goal of maximum profit conversion in a preset frequency range of requests; and
an optimal bid module for determining the first candidate bid as an optimal bid when there is a first candidate bid that can satisfy the maximum revenue conversion objective in the number of threshold requests;
wherein the first group of users are users who can see the experimental advertisement in the advertisement delivery platform, and the second group of users are users who can not see the experimental advertisement in the advertisement delivery platform;
and the maximum profit conversion target is determined according to the behavior data of the first group of users and the second group of users and the putting cost of the experimental advertisement.
9. A computer device, comprising: memory, processor and executable instructions stored in the memory and executable in the processor, characterized in that the processor implements the method according to any of claims 1-7 when executing the executable instructions.
10. A computer-readable storage medium having stored thereon computer-executable instructions, which when executed by a processor, implement the method of any one of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910823275.8A CN112446724A (en) | 2019-09-02 | 2019-09-02 | Bidding method, device and equipment based on effect evaluation and readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910823275.8A CN112446724A (en) | 2019-09-02 | 2019-09-02 | Bidding method, device and equipment based on effect evaluation and readable storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112446724A true CN112446724A (en) | 2021-03-05 |
Family
ID=74734290
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910823275.8A Pending CN112446724A (en) | 2019-09-02 | 2019-09-02 | Bidding method, device and equipment based on effect evaluation and readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112446724A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114186050A (en) * | 2021-12-06 | 2022-03-15 | 北京达佳互联信息技术有限公司 | Resource recommendation method and device, electronic equipment and storage medium |
-
2019
- 2019-09-02 CN CN201910823275.8A patent/CN112446724A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114186050A (en) * | 2021-12-06 | 2022-03-15 | 北京达佳互联信息技术有限公司 | Resource recommendation method and device, electronic equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US8583487B2 (en) | Method and system for media initialization via data sharing | |
| US20160132935A1 (en) | Systems, methods, and apparatus for flexible extension of an audience segment | |
| CN107845016B (en) | Information output method and device | |
| US20120284128A1 (en) | Order-independent approximation for order-dependent logic in display advertising | |
| CN111784400A (en) | Advertisement bidding method, device, electronic device and computer readable medium | |
| CN111210255B (en) | Advertisement pushing method and device and electronic equipment | |
| CN110889725B (en) | Online advertisement CTR estimation method, device, equipment and storage medium | |
| US10997634B2 (en) | Methods for determining targeting parameters and bids for online ad distribution | |
| US20210374809A1 (en) | Artificial intelligence techniques for bid optimization used for generating dynamic online content | |
| CN112446764A (en) | Game commodity recommendation method and device and electronic equipment | |
| CN104967690A (en) | Information push method and device | |
| CN108985810B (en) | Method and device for advertising on demand side platform | |
| CN114331543A (en) | Advertisement propagation method for large-scale crowd orientation and dynamic scene matching | |
| CN112232846B (en) | Method, device, medium and equipment for determining characteristic value based on multimedia file | |
| CN111798261A (en) | Information updating method and device | |
| CN108665312B (en) | Method and apparatus for generating information | |
| CN112446724A (en) | Bidding method, device and equipment based on effect evaluation and readable storage medium | |
| CN102498498A (en) | Expressive bidding in online advertising auctions | |
| CN107527128B (en) | Resource parameter determination method and equipment for advertisement platform | |
| CN109961161B (en) | Commodity management method, system, electronic device and computer readable medium | |
| CN112308648A (en) | Information processing method and device | |
| CN116167803A (en) | Advertisement putting method and device based on signaling data | |
| CN115936764A (en) | Product promotion method and device | |
| CN110838019A (en) | Method and device for determining trial supply distribution crowd | |
| CN109033343B (en) | Method and apparatus for generating information |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |