Advertisement data processing method and device, electronic equipment, storage medium and product
By fitting the probability distribution of bids and premiums from the demand side of advertising, the system automatically determines the minimum price for advertising, solving the problems of low efficiency in setting minimum prices for advertising and the risk of failed auctions, thus achieving a more efficient and reasonable setting of minimum prices for advertising.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- HANGZHOU NETEASE CLOUD MUSIC TECH CO LTD
- Filing Date
- 2025-12-17
- Publication Date
- 2026-05-12
AI Technical Summary
Current technologies have low efficiency in configuring minimum advertising bids, and there is a risk that advertisements may fail to sell due to unreasonable manual configuration.
By acquiring historical bidding data from advertising demanders, fitting the probability distribution of bids and premiums, evaluating the pricing value of candidate ad floor prices, and automating the decision on the target ad floor price.
It improves the efficiency of setting the minimum price for advertising, reduces the risk of advertising failing to sell, and enables reasonable evaluation and setting of the minimum price for advertising, as well as analysis of the patterns of bid and premium changes.
Smart Images
Figure CN122022927A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of data processing technology, and specifically to advertising data processing methods, apparatus, electronic devices, storage media, and products. Background Technology
[0002] In the digital advertising field, media outlets (such as website owners and app developers) can provide advertising services to demanders and drive business growth by selling ad space, while the media outlets also receive advertising revenue, creating a win-win situation for both parties.
[0003] Digital advertising can be sold through auction-based bidding. Buyers bid for each ad impression, and media outlets sell the ads based on a "highest bidder wins" principle. Typically, to prevent ads from being "sold cheaply," media outlets include a reserve price in each auction. This informs bidders that bids cannot be lower than the reserve price; only bids above the reserve price have a chance of winning. While this reserve price mechanism protects media outlets from being "sold cheaply," it can also lead to the risk of no bids being placed. For example, if the reserve price is set too high, no buyer may bid higher.
[0004] In related technologies, the minimum price for advertising is usually set manually based on human experience. However, with the increasing demand for competitive bidding advertising, manually setting the minimum price is inefficient and can easily lead to the risk of the advertisement failing to sell due to an unreasonable minimum price. Summary of the Invention
[0005] This invention provides an advertising data processing method, apparatus, electronic device, storage medium, and product, aiming to improve the low efficiency of advertising reserve price configuration in related technologies and the risk of advertising auction failure due to unreasonable advertising reserve price configuration.
[0006] In a first aspect, the present invention provides an advertising data processing method, the advertising data processing method comprising: Obtain the historical advertising bidding data corresponding to the floor price of each candidate advertisement. The historical advertising bidding data includes the bidding data and premium data of each advertising demander for the bidding advertisement with the floor price of the candidate advertisement. For each candidate ad floor price, the bid probability distribution of the candidate ad floor price is fitted with the bid data corresponding to the candidate ad floor price, and the premium probability distribution of the candidate ad floor price is fitted with the premium data corresponding to the candidate ad floor price. The pricing value of the candidate ad floor price is evaluated based on the fitted bid probability distribution and the fitted premium probability distribution. The target ad floor price is determined from the candidate ad floor prices based on their pricing value.
[0007] In some optional implementations, the step of fitting the probability distribution of the bids for the candidate ad floor prices to the bid data corresponding to the candidate ad floor prices includes: Whether each advertising demander bids is used as a random variable in the bidding probability distribution, and the parameters of the bidding probability distribution are estimated based on the bidding data corresponding to the candidate advertising floor price.
[0008] In some alternative implementations, the bid probability distribution is a discrete probability distribution.
[0009] In some optional implementations, the step of estimating the parameters of the bid probability distribution based on the bid data corresponding to the candidate ad floor price includes: Based on the bid data corresponding to the candidate ad floor price, the parameters of the bid probability distribution are estimated using the conjugate prior method.
[0010] In some optional implementations, the step of fitting the probability distribution of the premium of the candidate ad floor price to the premium data corresponding to the candidate ad floor price includes: The premium of each advertising demander is used as the random variable of the premium probability distribution, and the parameters of the premium probability distribution are estimated based on the premium data corresponding to the candidate advertising base price.
[0011] In some alternative implementations, the premium probability distribution is a continuous probability distribution.
[0012] In some optional implementations, the step of estimating the parameters of the premium probability distribution based on the premium data corresponding to the candidate ad floor price includes: Based on the premium data corresponding to the candidate advertisement's base price, the parameters of the premium probability distribution are estimated using the conjugate prior method.
[0013] In some optional implementations, evaluating the pricing value of the candidate ad floor price based on the fitted bid probability distribution and the fitted premium probability distribution includes: Based on the fitted bid probability distribution, determine the first expected value of the bid probability distribution; Based on the fitted premium probability distribution, determine the second expected value of the premium probability distribution; The pricing value of the candidate ad floor price is determined based on the first expected value, the second expected value, and the candidate ad floor price.
[0014] In some alternative implementations, the pricing value of the candidate ad floor price is the product of the second expected value and the candidate ad floor price, and the first expected value.
[0015] In some optional implementations, determining the target ad floor price from the candidate ad floor prices based on the pricing value of each candidate ad floor price includes: The candidate ad with the highest pricing value among all candidate ad prices is determined as the target ad price.
[0016] In a second aspect, the present invention provides an advertising data processing apparatus, the advertising data processing apparatus comprising: The acquisition module is used to acquire the historical advertising bidding data corresponding to the base price of each candidate advertisement. The historical advertising bidding data includes the bidding data and premium data of each advertising demander for the bidding advertisement with the base price of the candidate advertisement. The fitting module is used to fit the probability distribution of the bid probability of each candidate ad base price to the bid data corresponding to the candidate ad base price, and to fit the probability distribution of the premium probability of the candidate ad base price to the premium data corresponding to the candidate ad base price. The base price evaluation module is used to evaluate the pricing value of the candidate advertisement base price based on the fitted bid probability distribution and the fitted premium probability distribution. The base price determination module determines the target ad base price from among the candidate ad base prices based on the pricing value of each candidate ad base price.
[0017] Thirdly, the present invention provides an electronic device, comprising: a memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the computer instructions to perform the advertising data processing method of the first aspect or any corresponding embodiment described above.
[0018] Fourthly, the present invention provides a computer-readable storage medium storing computer instructions for causing a computer to perform the advertising data processing method of the first aspect or any corresponding embodiment described above.
[0019] Fifthly, the present invention provides a computer program product, including computer instructions for causing a computer to execute the advertising data processing method of the first aspect or any corresponding embodiment thereof.
[0020] According to the advertising data processing method, apparatus, electronic device, storage medium, and product of the present invention, the bidding probability distribution and premium probability distribution of advertising demanders for each candidate advertising reserve price are fitted by historical bidding data and premium data of advertising demanders to evaluate the pricing value of each candidate advertising reserve price. Based on the pricing value of each candidate advertising reserve price, the target advertising reserve price for the advertising space to be auctioned is determined, thereby realizing automated configuration of advertising reserve prices without manual setting, improving the configuration efficiency of advertising reserve prices. Moreover, by fitting the bidding probability distribution and premium probability distribution of advertising demanders for each candidate advertising reserve price, it is beneficial to analyze and perceive the bidding change pattern and premium change pattern of advertising demanders, thereby facilitating the reasonable evaluation of the pricing value of candidate advertising reserve prices, and thus enabling the determination of a reasonable target advertising reserve price, which helps to mitigate the risk of advertising auction failure due to unreasonable configuration of advertising reserve prices. Attached Figure Description
[0021] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.
[0022] Figure 1 This is a schematic diagram illustrating an application scenario of an embodiment of the present invention; Figure 2 A flowchart illustrating an advertising data processing method provided in an embodiment of the present invention; Figure 3 This is a flowchart illustrating a method for evaluating the minimum price of a candidate advertisement in an embodiment of the present invention; Figure 4 This is a schematic diagram illustrating an application scenario of an advertising data processing method according to an embodiment of the present invention; Figure 5 A structural block diagram of an advertising data processing device provided in an embodiment of the present invention; Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation
[0023] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0024] It is understood that before using the technical solutions disclosed in the various embodiments of the present invention, users should be informed of the types, scope of use, and usage scenarios of the personal information involved in the present invention and their authorization should be obtained in accordance with relevant laws and regulations through appropriate means.
[0025] The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.
[0026] Figure 1 This is a schematic diagram of one application scenario of the present invention. As an optional application scenario of the present invention, such as... Figure 1 As shown, the advertising bidding system includes an Ad Exchange (ADX), a Sell-Side Platform (SSP), and a Demand-Side Platform (DSP).
[0027] The Ad Exchange (ADX) is a core hub in the programmatic advertising ecosystem. It connects Demand-Side Platforms (DSPs) and Service Provider Platforms (SSPs) through Real-Time Bidding (RTB) technology to enable automated trading and precise matching of advertising resources.
[0028] Demand-side platforms (DSPs) connect advertisers, i.e., those who buy ad space; while media-side platforms (SSPs) connect advertisers, i.e., those who sell ad space.
[0029] When a user visits a media page, the advertising media platform (SSP) sends an advertising request to the advertising exchange (ADX), including ad placement information and user data. After receiving the advertising request from the SSP, the ADX simultaneously sends bidding requests to multiple advertising demand-side platforms (DSPs). The bidding requests can include the reserve price for the ad placement. Each DSP returns bidding data for the media's ad placement in real time according to the bidding strategy set by the advertiser. The ADX then uses a pre-defined algorithm to determine the winning bidder from among the multiple advertising demand-side platforms.
[0030] In related technologies, the setting of reserve prices for advertising auctions typically involves pre-configuring different reserve prices for different ad slots. This approach has a certain rationale, based on the premise that different ad slots have different commercial values, and different reserve prices are set according to their value.
[0031] For example, in most cases, the splash screen ad slots of media application apps have higher commercial value due to their larger size and stronger visual impact, and advertisers are more willing to pay high prices to purchase these splash screen ad slots. Conversely, banner ad slots within apps have relatively lower commercial value because of their smaller display area, less variety in styles, and difficulty in attracting attention.
[0032] Commercial value can be quantitatively measured using eCPM (Effective Cost Per Mille), which refers to the average cost advertisers are willing to pay for 1,000 ad impressions, or the average revenue media outlets can obtain for 1,000 ad impressions. The formula for calculating eCPM is: eCPM = Total Ad Revenue / Total Ad Impressions × 1000.
[0033] For media outlets, eCPM is the most direct indicator of the value of their ad placements. The higher the eCPM, the higher the quality of your traffic and the stronger your monetization ability. Media outlets usually multiply the eCPM of each ad placement by a coefficient β to determine the final base price for that ad placement: Base price = eCPM × β.
[0034] Among them, the coefficient β is usually set manually based on human experience and relies on manual adjustment. This method of manually setting the reserve price is difficult to meet the growing demand for bidding advertising. The efficiency of setting the reserve price for advertising is low, and there is a risk of advertising failing to sell due to unreasonable reserve prices set manually.
[0035] Therefore, embodiments of the present invention provide an advertising data processing method, apparatus, electronic device, storage medium, and product, which aim to effectively improve the low efficiency of advertising reserve price configuration and the risk of advertising auction failure due to unreasonable advertising reserve price configuration in the above-mentioned related technologies.
[0036] According to an embodiment of the present invention, an embodiment of an advertising data processing method is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.
[0037] This invention provides an advertising data processing method that can be used in the advertising trading platform or advertising media platform described above. Figure 2 This is a flowchart illustrating an advertising data processing method provided in an embodiment of the present invention, as shown below. Figure 2 As shown, the advertising data processing method includes the following steps.
[0038] Step S201: Obtain the historical advertising bidding data corresponding to the base price of each candidate advertisement. The historical advertising bidding data includes the bidding data and premium data of each advertising demander for the bidding advertisement with the base price of the candidate advertisement.
[0039] In this embodiment of the invention, multiple candidate ad reserve prices are pre-configured for the ad space to be auctioned. The candidate ad reserve prices are pre-set auction reserve prices that can be selected for the ad space to be auctioned. These multiple candidate ad reserve prices can also be selected based on the historical bidding experience data of the ad space to be auctioned.
[0040] For the ad slots to be auctioned, obtain their historical ad bidding data, and classify the historical ad bidding data according to the corresponding ad reserve price, thereby obtaining the historical ad bidding data corresponding to each candidate ad reserve price.
[0041] By analyzing the historical advertising bidding data corresponding to the reserve price of each candidate advertisement, we can obtain the bidding data and premium data of each advertising demander for the advertisement space with the reserve price of the candidate advertisement. The bidding data refers to the bidding price given by the advertising demander when participating in the bidding auction for the advertisement space with the reserve price of the candidate advertisement. The premium data refers to the difference between the bidding price given by the advertising demander when participating in the bidding auction for the advertisement space with the reserve price of the candidate advertisement and the reserve price of the candidate advertisement space.
[0042] Step S202: For each candidate ad base price, perform probability distribution fitting on the bid probability distribution of the candidate ad base price based on the bid data corresponding to the candidate ad base price, and perform probability distribution fitting on the premium probability distribution of the candidate ad base price based on the premium data corresponding to the candidate ad base price.
[0043] For each candidate ad's reserve price, the probability of whether an ad demander bids is represented by a bidding probability distribution. In step S202, the bidding probability distribution is fitted using the bidding data corresponding to the candidate ad's reserve price to solve for its distribution parameters. The bidding probability distribution describes the probability pattern of whether an ad demander bids for an ad slot with a reserve price equal to the candidate ad's reserve price.
[0044] The probability of premium bidding by the advertising demand side is represented by a premium probability distribution. In step S202, the premium probability distribution is fitted using the premium data corresponding to the candidate advertisement's base price to solve for the distribution parameters of the premium probability distribution. The premium probability distribution describes the probability pattern of premium bidding by the advertising demand side in bidding for an advertisement with a base price equal to the candidate advertisement's base price.
[0045] Step S203: Evaluate the pricing value of the candidate ad floor price based on the fitted bid probability distribution and the fitted premium probability distribution.
[0046] Probability distributions can reveal the inherent patterns and overall structure of data. In this embodiment of the invention, by analyzing the probability distribution of bids from advertising demanders for ad slots with a base price equal to the candidate ad base price, key characteristics such as the trend (e.g., mean), the dispersion (e.g., variance), and the range of bids from advertising demanders can be quickly analyzed. By analyzing the probability distribution of premiums from advertising demanders for ad slots with a base price equal to the candidate ad base price, key characteristics such as the premium patterns and premium levels from advertising demanders can be quickly analyzed.
[0047] Therefore, by fitting the bidding probability distribution and premium probability distribution of the advertising demand side for the ad space with the candidate ad floor price as the floor price, it is helpful to analyze and evaluate the rationality and value of setting the ad floor price of the ad space as the candidate ad floor price.
[0048] In step S203, the pricing value of the candidate ad floor price is analyzed and evaluated by fitting the bid probability distribution and the premium probability distribution. The pricing value refers to the value of setting the ad floor price of the ad slot as the candidate ad floor price. As a key indicator reflecting the rationality of the candidate ad floor price, the pricing value can also be used to analyze and evaluate the gains and risks brought about by setting the ad floor price of the ad slot as the candidate ad floor price. The lower the pricing value of the candidate ad floor price, the lower the bid probability and premium probability of each ad demander. Therefore, the greater the risk of the ad auction failing to sell when the ad floor price of the ad slot is set as the candidate ad floor price.
[0049] Step S204: Determine the target ad price from the candidate ad prices based on the pricing value of each candidate ad price.
[0050] In this embodiment of the invention, based on the assessed pricing value of each candidate advertisement's reserve price, the final auction reserve price of the advertisement space to be auctioned, i.e., the target advertisement reserve price, is determined.
[0051] According to the advertising data processing method provided in this embodiment of the invention, the bidding probability distribution and premium probability distribution of the advertising demander for each candidate advertising reserve price are fitted by historical bidding data and premium data of the advertising demander. This is to evaluate the pricing value of each candidate advertising reserve price, and to determine the target advertising reserve price for the advertising space to be auctioned based on the pricing value of each candidate advertising reserve price. This achieves automated configuration of advertising reserve prices without manual setting, improving the efficiency of advertising reserve price configuration. Moreover, by fitting the bidding probability distribution and premium probability distribution of the advertising demander for each candidate advertising reserve price, it is possible to analyze and perceive the bidding change pattern and premium change pattern of the advertising demander. This is conducive to reasonably evaluating the pricing value of the candidate advertising reserve prices, and thus to deciding on a reasonable target advertising reserve price. This helps to mitigate the risk of advertising auction failure due to unreasonable advertising reserve price configuration.
[0052] In some optional implementations, in step S202 above, the probability distribution fitting of the bid probability distribution of the candidate ad floor price based on the bid data corresponding to the candidate ad floor price may further include: using whether each ad demander bids as a random variable of the bid probability distribution, and estimating the parameters of the bid probability distribution based on the bid data corresponding to the candidate ad floor price.
[0053] In some optional implementations, in the bidding data corresponding to the candidate ad floor price, the bidding behavior and non-bidding behavior of the ad demander can be represented by 1 and 0, where 1 indicates that the ad demander bids and 0 indicates that the ad demander does not bid. Therefore, whether the ad demander bids is a discrete random variable, and the probability distribution of the ad demander's bid for the candidate ad floor price can be a discrete probability distribution.
[0054] In some alternative implementations, the bid probability distribution can be a Bernoulli distribution, denoted as IsBidding~Bernoulli(p), where IsBidding represents the random variable indicating whether the advertising demander will bid.
[0055] In some optional implementations, whether each advertising demander bids is used as a random variable in the bidding probability distribution. Based on the bidding data corresponding to the candidate advertising floor price, a preset parameter estimation algorithm is used to estimate the parameters of the bidding probability distribution to obtain the distribution parameters of the bidding probability distribution. For example, the bidding probability distribution is a Bernoulli distribution (p), and the distribution parameter of the demand solution is parameter p.
[0056] In this embodiment of the invention, no special restrictions are placed on the parameter estimation algorithm for the bid probability distribution. Any suitable parameter estimation algorithm can be used, such as maximum likelihood estimation (MLE), maximum a posteriori estimation (MAP), Bayesian method, conjugate prior method, etc.
[0057] In some optional implementations, the parameter estimation algorithm for the bid probability distribution employs the conjugate prior method. The above-mentioned parameter estimation of the bid probability distribution based on the bid data corresponding to the candidate ad floor price may further include: estimating the parameters of the bid probability distribution using the conjugate prior method based on the bid data corresponding to the candidate ad floor price, so as to solve for the distribution parameters of the bid probability distribution.
[0058] Taking the Bernoulli distribution as an example, the conjugate prior method is used to solve for the distribution parameters of the Bernoulli distribution. The conjugate prior of the Bernoulli distribution is the beta distribution. The distribution parameter p of the Bernoulli distribution follows the beta distribution, denoted as p ~ Beta(α, β). α can be used to represent the count of bids made by the advertising demand side, and β can be used to represent the count of no bids made by the advertising demand side.
[0059] The process of solving the distribution parameters of the Bernoulli distribution using the conjugate prior method is briefly described below.
[0060] First, initialize the hyperparameters α and β of the beta distribution. For example, let α=1 and β=1, with a uniform prior distribution, indicating no prior information.
[0061] Secondly, based on the bidding data corresponding to the candidate ad floor price, the number of times each ad demander bids and does not bid is counted, and a likelihood function is constructed based on the number of times bids and does not bid.
[0062] Subsequently, using the prior parameters of the prior distribution (i.e., the beta distribution) and the parameters of the likelihood function, the hyperparameters of the posterior distribution are obtained. According to the Bayesian statistical principle, the posterior distribution is proportional to the product of the prior distribution and the likelihood function, thus constructing the posterior distribution.
[0063] Finally, the distribution parameters of the Bernoulli distribution are estimated based on the mean of the posterior distribution, and the distribution parameters p of the Bernoulli distribution are obtained by solving the problem.
[0064] In some optional implementations, in step S202 above, fitting the probability distribution of the premium probability distribution of the candidate ad floor price based on the premium data corresponding to the candidate ad floor price may further include: using the premium of each ad demander as a random variable of the premium probability distribution, and estimating the parameters of the premium probability distribution based on the premium data corresponding to the candidate ad floor price.
[0065] In some alternative implementations, the premium data corresponding to the candidate ad floor price is a positive number greater than zero. Ad demanders usually expect to obtain ad display opportunities at a lower cost, so the premium of ad demanders is usually low, and its mean is usually around 0. Moreover, the premium of each ad demander is usually concentrated in a certain range, and has a long tail in other ranges. Therefore, the premium of ad demanders can be regarded as a continuous random variable, and the probability distribution of the premium of ad demanders for the candidate ad floor price can adopt a continuous probability distribution.
[0066] In some alternative implementations, the premium probability distribution can be a gamma distribution. , denoted as: , A random variable representing the premium of the advertising demand side.
[0067] In some optional implementations, the premium of each advertising demander is used as the random variable of the premium probability distribution. Based on the premium data corresponding to the candidate advertising floor price, a preset parameter estimation algorithm is used to estimate the parameters of the premium probability distribution to obtain the distribution parameters of the premium probability distribution. For example, the premium probability distribution is a gamma distribution. The distribution parameter of the required solution is parameter θ.
[0068] In this embodiment of the invention, no special restrictions are placed on the parameter estimation algorithm for the premium probability distribution. Any suitable parameter estimation algorithm can be used, such as maximum likelihood estimation (MLE), maximum a posteriori estimation (MAP), Bayesian method, conjugate prior method, etc.
[0069] In some optional implementations, the parameter estimation algorithm for the premium probability distribution adopts the conjugate prior method. The above-mentioned parameter estimation of the premium probability distribution based on the premium data corresponding to the candidate ad floor price may further include: estimating the parameters of the premium probability distribution based on the premium data corresponding to the candidate ad floor price using the conjugate prior method, so as to solve for the distribution parameters of the premium probability distribution.
[0070] Taking the premium probability distribution as a gamma distribution as an example, the distribution parameters of the gamma distribution are solved using the conjugate prior method. The conjugate prior of the gamma distribution is the inverse gamma distribution. A likelihood function is constructed using the premium data corresponding to the candidate advertising base price. The hyperparameters of the posterior distribution are obtained using the prior parameters of the prior distribution (i.e., the inverse gamma distribution) and the parameters of the likelihood function. According to Bayesian statistical principles, the posterior distribution is proportional to the product of the prior distribution and the likelihood function. Thus, the posterior distribution is constructed. Finally, the distribution parameters of the gamma distribution are estimated based on the mean of the posterior distribution, and the distribution parameters of the gamma distribution are obtained.
[0071] Figure 3This is a flowchart illustrating a method for evaluating the reserve price of candidate advertisements in an embodiment of the present invention. In some optional embodiments, such as... Figure 3 As shown, in step S203 above, evaluating the pricing value of the candidate ad floor price based on the fitted bid probability distribution and the fitted premium probability distribution may further include: Step S301: Determine the first expected value of the bid probability distribution based on the fitted bid probability distribution.
[0072] The first expected value of the bid probability distribution is calculated based on the distribution parameters obtained from parameter estimation of the bid probability distribution. Taking the bid probability distribution as a Bernoulli distribution as an example, the first expected value of the Bernoulli distribution is... , where p is the distribution parameter of the Bernoulli distribution.
[0073] Step S302: Determine the second expected value of the premium probability distribution based on the fitted premium probability distribution.
[0074] The second expected value of the premium probability distribution is calculated based on the distribution parameters obtained from parameter estimation of the premium probability distribution. Taking the gamma distribution as an example, the second expected value of the gamma distribution is... , where k and θ are the distribution parameters of the gamma distribution.
[0075] Step S303: Determine the pricing value of the candidate ad's floor price based on the first expected value, the second expected value, and the candidate ad's floor price.
[0076] In some alternative implementations, the pricing value of the candidate ad floor price is the product of the second expected value and the candidate ad floor price, and the first expected value.
[0077] For example, taking a Bernoulli distribution for the bid probability distribution and a gamma distribution for the premium probability distribution, the first expected value is... The second expected value is The pricing value of the candidate ad floor price is expressed as: eCPM represents the pricing value of the candidate ad floor price, and FloorPrice represents the candidate ad floor price.
[0078] In some optional implementations, in step S205 above, determining the target ad price from the candidate ad prices based on the pricing value of each candidate ad price may further include: determining the candidate ad price with the largest pricing value among the candidate ad prices as the target ad price.
[0079] In some alternative implementations, the pricing value of the candidate ad floor price can be used to assess the expected revenue that the media outlet can generate when the ad floor price of the ad slot is set as the candidate ad floor price. The lower the bidding probability and premium probability of each ad demander, the lower the expected revenue for the media outlet, and vice versa.
[0080] Figure 4 This is a schematic diagram illustrating an application scenario of an advertising data processing method according to an embodiment of the present invention. In some application scenarios, such as... Figure 4 As shown, steps S201 to S202 can be performed offline, while steps S203 to S204 can be performed online.
[0081] After the advertising auction ends, the historical advertising bidding data of each advertising demand party (such as demand party A, B, and C), including each advertising demand party's participation in the advertising auction, bidding data, and premium data, will be recorded and stored in the database in real time.
[0082] In the offline phase, historical ad bidding data for each ad demander is retrieved from the database. Based on this data, the conjugate prior method is used to fit the bid probability distribution and premium probability distribution for each ad demander at each candidate ad's base price. This yields the distribution parameters for the bid probability distribution and the premium probability distribution, which are then stored in a distribution parameter database. The database is then updated with the latest distribution parameters for both the bid probability and premium probability distributions. For example, the bid probability distribution might be a Bernoulli distribution with parameter p, and the premium probability distribution might be a gamma distribution with parameters k and θ.
[0083] When a user visits the media provider's application (APP), the ad display mechanism is triggered in response to the user's access, and the user enters the online phase.
[0084] During the online phase, the media platform retrieves the latest distribution parameters of the bidding probability distribution and the premium probability distribution from the distribution parameter library. Based on these latest distribution parameters, it evaluates and calculates the pricing value of each candidate ad's reserve price for each ad demander. Then, it determines the target ad reserve price from the pricing value of each candidate ad's reserve price and uses it as the auction reserve price displayed in this ad auction. It then sends ad requests carrying the target ad reserve price to each ad demander. Each ad demander (such as demanders A, B, and C) participates in the bidding according to their respective advertising needs. After the ad auction ends, the historical ad bidding data of each ad demander is stored in the database.
[0085] This invention also provides an advertising data processing apparatus for implementing the above embodiments and preferred embodiments; details already described will not be repeated. As used below, the term "module" can refer to a combination of software and / or hardware that performs a predetermined function. Although the apparatus described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.
[0086] This invention provides an advertising data processing device. Figure 5 A structural block diagram of an advertising data processing device provided in an embodiment of the present invention is shown below. Figure 5 As shown, the advertising data processing device includes: The acquisition module 501 is used to acquire the historical advertising bidding data corresponding to the base price of each candidate advertisement. The historical advertising bidding data includes the bidding data and premium data of each advertising demander for the bidding advertisement with the base price of the candidate advertisement.
[0087] The fitting module 502 is used to fit the probability distribution of the bid probability distribution of the candidate ad floor price based on the bid data corresponding to the candidate ad floor price for each candidate ad floor price, and to fit the probability distribution of the premium probability distribution of the candidate ad floor price based on the premium data corresponding to the candidate ad floor price.
[0088] The base price assessment module 503 is used to assess the pricing value of candidate ad base prices based on the fitted bid probability distribution and the fitted premium probability distribution.
[0089] The base price determination module 504 is used to determine the target ad base price from the base prices of each candidate ad based on the pricing value of each candidate ad base price.
[0090] The advertising data processing apparatus provided in this embodiment of the invention can execute the advertising data processing method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects for executing the method. Further functional descriptions of the various modules and units described above are the same as in the corresponding embodiments described above, and will not be repeated here.
[0091] Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention.
[0092] The following is a detailed reference. Figure 6This diagram illustrates a structural schematic suitable for implementing an electronic device according to embodiments of the present invention. The electronic device may include a processor (e.g., a central processing unit, graphics processor, etc.) 601, which can perform various appropriate actions and processes according to a program stored in read-only memory (ROM) 602 or a program loaded from memory 608 into random access memory (RAM) 603. RAM 603 also stores various programs and data required for the operation of the electronic device. The processor 601, ROM 602, and RAM 603 are interconnected via bus 604. Input / output (I / O) interface 605 is also connected to bus 604.
[0093] Typically, the following devices can be connected to I / O interface 605: input devices 606 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 607 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; memory devices 608 including, for example, magnetic tapes, hard disks, etc.; and communication devices 609. Communication device 609 allows electronic devices to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 6 Electronic devices with various devices are shown, but it should be understood that it is not required to implement or have all of the devices shown, and more or fewer devices may be implemented or have instead.
[0094] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device 609, or installed from a memory 608, or installed from a ROM 602. When the computer program is executed by the processor 601, it performs the functions defined in the advertising data processing method of the embodiments of the present invention.
[0095] Figure 6 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of use of the embodiments of the present invention.
[0096] This invention also provides a computer-readable storage medium. The methods described above according to embodiments of the invention can be implemented in hardware or firmware, or implemented as computer code that can be recorded on a storage medium, or implemented as computer code downloaded via a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and then stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code. When the software or computer code is accessed and executed by the computer, processor, or hardware, the advertising data processing method shown in the above embodiments is implemented.
[0097] A portion of this invention can be applied as a computer program product, such as computer program instructions, which, when executed by a computer, can invoke or provide the methods and / or technical solutions according to the invention through the operation of the computer. Those skilled in the art will understand that the forms in which computer program instructions exist in a computer-readable medium include, but are not limited to, source files, executable files, installation package files, etc. Correspondingly, the ways in which computer program instructions are executed by a computer include, but are not limited to: the computer directly executing the instructions, or the computer compiling the instructions and then executing the corresponding compiled program, or the computer reading and executing the instructions, or the computer reading and installing the instructions and then executing the corresponding installed program. Here, the computer-readable medium can be any available computer-readable storage medium or communication medium accessible to a computer.
[0098] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and such modifications and variations all fall within the scope defined by the appended claims.
Claims
1. A method for processing advertising data, characterized in that, The method includes: Obtain the historical advertising bidding data corresponding to the floor price of each candidate advertisement. The historical advertising bidding data includes the bidding data and premium data of each advertising demander for the bidding advertisement with the floor price of the candidate advertisement. For each candidate ad floor price, the bid probability distribution of the candidate ad floor price is fitted with the bid data corresponding to the candidate ad floor price, and the premium probability distribution of the candidate ad floor price is fitted with the premium data corresponding to the candidate ad floor price. The pricing value of the candidate ad floor price is evaluated based on the fitted bid probability distribution and the fitted premium probability distribution. The target ad floor price is determined from the candidate ad floor prices based on their pricing value.
2. The method according to claim 1, characterized in that, The step of fitting the probability distribution of the bid probability distribution of the candidate ad floor price based on the bid data corresponding to the candidate ad floor price includes: Whether each advertising demander bids is used as a random variable in the bidding probability distribution, and the parameters of the bidding probability distribution are estimated based on the bidding data corresponding to the candidate advertising floor price.
3. The method according to claim 2, characterized in that, The step of estimating the parameters of the bid probability distribution based on the bid data corresponding to the candidate ad floor price includes: Based on the bid data corresponding to the candidate ad floor price, the parameters of the bid probability distribution are estimated using the conjugate prior method.
4. The method according to claim 1, characterized in that, The step of fitting the probability distribution of the premium probability distribution of the candidate ad floor price based on the premium data corresponding to the candidate ad floor price includes: The premium of each advertising demander is used as the random variable of the premium probability distribution, and the parameters of the premium probability distribution are estimated based on the premium data corresponding to the candidate advertising base price.
5. The method according to claim 4, characterized in that, The step of estimating the parameters of the premium probability distribution based on the premium data corresponding to the candidate ad floor price includes: Based on the premium data corresponding to the candidate advertisement's base price, the parameters of the premium probability distribution are estimated using the conjugate prior method.
6. The method according to claim 1, characterized in that, The step of evaluating the pricing value of the candidate ad floor price based on the fitted bid probability distribution and the fitted premium probability distribution includes: Based on the fitted bid probability distribution, determine the first expected value of the bid probability distribution; Based on the fitted premium probability distribution, determine the second expected value of the premium probability distribution; The pricing value of the candidate ad floor price is determined based on the first expected value, the second expected value, and the candidate ad floor price.
7. An advertising data processing device, characterized in that, The device includes: The acquisition module is used to acquire the historical advertising bidding data corresponding to the base price of each candidate advertisement. The historical advertising bidding data includes the bidding data and premium data of each advertising demander for the bidding advertisement with the base price of the candidate advertisement. The fitting module is used to fit the probability distribution of the bid probability of each candidate ad base price to the bid data corresponding to the candidate ad base price, and to fit the probability distribution of the premium probability of the candidate ad base price to the premium data corresponding to the candidate ad base price. The base price evaluation module is used to evaluate the pricing value of the candidate advertisement base price based on the fitted bid probability distribution and the fitted premium probability distribution. The base price determination module determines the target ad base price from among the candidate ad base prices based on the pricing value of each candidate ad base price.
8. An electronic device, characterized in that, include: A memory and a processor are communicatively connected, the memory storing computer instructions, and the processor executing the computer instructions to perform the advertising data processing method according to any one of claims 1 to 6.
9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing the computer to perform the advertising data processing method according to any one of claims 1 to 6.
10. A computer program product, characterized in that, Includes computer instructions for causing a computer to perform the advertising data processing method according to any one of claims 1 to 6.