Advertisement processing method, related device and medium

Abstract

The invention provides an advertisement processing method, a related device and a medium. The method comprises the steps of determining predicted associated resource data of a target advertisement based on target associated resource data and predetermined resource constraint data of the target advertisement, and determining an initial bid of the target advertisement based on the predicted associated resource data; performing bid exploration based on the initial bid, and determining actual consumption resource data; if the actual consumption resource data meets the predetermined condition, determining expected consumption resource data based on the traffic consumption data and predetermined resource constraint data; and updating the initial bid based on comparison of the actual consumption resource data and the expected consumption resource data, and returning to the step of determining the expected consumption resource data of the target advertisement based on the flow consumption data and the predetermined resource constraint data until a predetermined exploration condition is reached, and determining a target bid of the target advertisement. The bidding accuracy of the advertisement can be improved. The method can be applied to various scenes such as big data, content promotion and cloud technology.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to the technical field of big data, and in particular, to an advertisement processing method, related apparatus and medium. BACKGROUND

[0002] Auction advertising is a new form of Internet advertising in which advertisers compete for advertising exposure by bidding for advertising opportunities. For advertisers, each advertising exposure opportunity can lead to subsequent conversions, and advertisers can obtain advertising exposure by providing more interactive resources, thereby bringing more advertising clicks, App activations, user order quantities, and more transaction opportunities. Based on this, the interactive resources provided by advertisers for advertising are particularly important for successful bidding and maximizing the conversion volume of advertising.

[0003] In related technologies, a bidding mode mainly based on CPM (cost per thousand impressions) and CPC (cost per click) is often used, in which advertisers bid for the number of advertising displays or the actual click behavior of users. Alternatively, an oCPA bidding mode is used, in which the amount of interactive resources provided in the bidding for advertising opportunities is manually set. Both of these modes have the problem of inaccurate bidding for advertising, which makes it difficult to improve the delivery effect of advertising. SUMMARY

[0004] The embodiments of the present disclosure provide an advertisement processing method, related apparatus and medium, which can improve the accuracy of bidding for advertising and thereby improve the delivery effect of advertising.

[0005] According to an aspect of the present disclosure, an advertisement processing method is provided, which includes:

[0006] obtaining target associated resource data of a target advertisement, wherein the target associated resource data is used to indicate the amount of input resources associated with the target advertisement in a target dimension;

[0007] determining predicted associated resource data of the target advertisement based on the target associated resource data and predetermined resource constraint data of the target advertisement, and determining an initial bid of the target advertisement based on the predicted associated resource data;

[0008] performing bid exploration based on the initial bid, and determining actual consumed resource data of the target advertisement;

[0009] if it is determined that the actual consumed resource data meets a predetermined condition, determining expected consumed resource data of the target advertisement based on previously obtained traffic consumption data and the predetermined resource constraint data;

[0010] updating the initial bid based on comparison between the actual consumed resource data and the expected consumed resource data, and returning to the step of determining the expected consumed resource data of the target advertisement based on preset traffic consumption data and the predetermined resource constraint data until the target advertisement reaches a predetermined exploration condition, and determining the target bid of the target advertisement based on the updated initial bid.

[0011] According to an aspect of the present disclosure, an advertisement processing apparatus is provided, which comprises:

[0012] an obtaining unit configured to obtain target associated resource data of a target advertisement, wherein the target associated resource data is used to indicate an amount of input resource associated with the target advertisement in a target dimension;

[0013] a first determining unit configured to determine predicted associated resource data of the target advertisement based on the target associated resource data and predetermined resource constraint data of the target advertisement, and determine an initial bid of the target advertisement based on the predicted associated resource data;

[0014] a second determining unit configured to perform bid exploration based on the initial bid, and determine actual consumed resource data of the target advertisement;

[0015] a third determining unit configured to, if it is determined that the actual consumed resource data meets a predetermined condition, determine expected consumed resource data of the target advertisement based on preset traffic consumption data and the predetermined resource constraint data;

[0016] an updating unit configured to update the initial bid based on comparison between the actual consumed resource data and the expected consumed resource data, and return to the step of determining the expected consumed resource data of the target advertisement based on preset traffic consumption data and the predetermined resource constraint data until the target advertisement reaches a predetermined exploration condition, and determine the target bid of the target advertisement based on the updated initial bid.

[0017] Optionally, the obtaining unit is configured to:

[0018] obtain target exposure data of the target advertisement in a target dimension;

[0019] if it is determined that the target exposure data does not reach a predetermined exposure threshold, obtain reference exposure data of the target advertisement in a higher level dimension of the target dimension;

[0020] replacing the target dimension with a parent dimension of the target dimension and replacing the target exposure data with the reference exposure data, and returning to the step of obtaining reference exposure data of the target advertisement in a parent dimension of the target dimension if it is determined that the target exposure data does not reach a predetermined exposure threshold, until the target exposure data reaches the predetermined exposure threshold, obtaining the input resource average value associated with the target dimension of the target advertisement;

[0021] determining the target associated resource data based on the input resource average value.

[0022] Optionally, the first determining unit is configured to:

[0023] determining the number of associated items generated by the target advertisement in a target time day in the target dimension;

[0024] determining the predicted associated resource data based on the number of associated items, the target associated resource data, and the predetermined resource constraint data.

[0025] Optionally, the determining the predicted associated resource data based on the number of associated items, the target associated resource data, and the predetermined resource constraint data comprises:

[0026] determining a first product based on a product of the number of associated items and the target associated resource data;

[0027] performing logarithmic scaling on a comparison result of the predetermined resource constraint data and the first product to obtain a logarithmic scaling result;

[0028] performing scaling processing on the logarithmic scaling result and 1 and the target associated resource data to obtain the predicted associated resource data.

[0029] Optionally, the first determining unit is configured to:

[0030] determining an exploration factor of the target advertisement based on a launch speed of the target advertisement and associated input resources;

[0031] determining the initial bid based on a product of an inverse of the exploration factor and the predicted associated resource data.

[0032] Optionally, the third determining unit is configured to:

[0033] performing fitting processing on the traffic consumption data to obtain a first curve fitting result indicating a traffic consumption distribution of a target time day and a second curve fitting result indicating a traffic consumption distribution of a non-target time day;

[0034] determining a current time day and cumulative time data of the current time day.

[0035] determine a target curve fitting result from the first curve fitting result and the second curve fitting result based on the current time day;

[0036] determine predicted traffic consumption data based on the cumulative time data and the target curve fitting result;

[0037] determine expected consumption resource data of the target advertisement based on a product of the predicted traffic consumption data and the predetermined resource constraint data.

[0038] Optionally, the fitting processing of the traffic consumption data to obtain the first curve fitting result indicating traffic consumption distribution of a target time day and the second curve fitting result indicating traffic consumption distribution of a non-target time day comprises:

[0039] determine a first proportion of cumulative traffic consumption amount of each sampling time point of the target time day in total traffic consumption amount of the target time day and a second proportion of cumulative traffic consumption amount of each sampling time point of the non-target time day in total traffic consumption amount of the non-target time day based on the traffic consumption data;

[0040] determine the first curve fitting result based on the first proportion of each sampling time point of the target time day and the least square method;

[0041] determine the second curve fitting result based on the second proportion of each sampling time point of the non-target time day and the least square method.

[0042] Optionally, the target curve fitting result comprises a plurality of candidate time intervals and a curve equation corresponding to each candidate time interval;

[0043] The determining of the predicted traffic consumption data based on the cumulative time data and the target curve fitting result comprises:

[0044] determine a target time interval from the plurality of candidate time intervals based on the cumulative time data;

[0045] substitute the cumulative time data into the curve equation corresponding to the target time interval to obtain the predicted traffic consumption data.

[0046] Optionally, the updating unit comprises:

[0047] a comparison module configured to determine a bid adjustment parameter of the target advertisement based on a comparison between the actual consumption resource data and the expected consumption resource data;

[0048] an updating module configured to update the initial bid based on the bid adjustment parameter.

[0049] Optionally, the comparing module is configured to:

[0050] determine a first value based on a quotient of the expected consumed resource data and the actual consumed resource data;

[0051] determine, between the first value and a first predetermined value, a value with a smaller numerical value as an intermediate parameter;

[0052] determine, between the intermediate parameter and a second predetermined value, a value with a larger numerical value as the bid adjustment parameter, wherein the first predetermined value is larger than the second predetermined value.

[0053] Optionally, the updating unit is configured to:

[0054] determine an upper limit parameter of a bid of the target advertisement;

[0055] determine a maximum bid of the target advertisement based on a product of the upper limit parameter of the bid and the predicted associated resource data;

[0056] determine, between the updated initial bid and the maximum bid, a value with a smaller numerical value as a target bid of the target advertisement.

[0057] Optionally, the advertisement processing apparatus further comprises a judging unit configured to:

[0058] determine whether the actual consumed resource data is non-zero data;

[0059] if it is determined that the actual consumed resource data is non-zero data, determine that the actual consumed resource data satisfies a predetermined condition;

[0060] if it is determined that the actual consumed resource data is not non-zero data, adjust the initial bid based on a predetermined bid adjustment factor until the actual consumed resource data is non-zero data, and replace the initial bid with an adjusted initial bid.

[0061] Optionally, the adjusting the initial bid based on the predetermined bid adjustment factor comprises:

[0062] determining a second sum of the predetermined bid adjustment factor and 1;

[0063] determining the adjusted initial bid based on a product of the second sum and the initial bid.

[0064] Optionally, the predetermined bid adjustment factor is determined by:

[0065] determining a winning difficulty level and a delivery scenario of the target advertisement;

[0066] determining a first score based on the winning difficulty level;

[0067] determining a second score based on the delivery scenario;

[0068] determining the predetermined bid adjustment factor based on the first score and the second score.

[0069] According to an aspect of the present disclosure, an electronic device is provided, which includes a memory storing a computer program and a processor implementing the advertisement processing method as described above when executing the computer program.

[0070] According to an aspect of the present disclosure, a computer readable storage medium is provided, which stores a computer program, the computer program being executed by a processor to implement the advertisement processing method as described above.

[0071] According to an aspect of the present disclosure, a computer program product is provided, which includes a computer program, the computer program being read and executed by a processor of a computer device, so that the computer device performs the advertisement processing method as described above.

[0072] In the embodiments of the present disclosure, in order to solve the problem that the inaccurate bidding of the target advertisement is caused by the manual setting of the bidding price of the target advertisement in the advertisement opportunity bidding process in the prior art, the bidding price of the target advertisement is dynamically adjusted according to the input resource condition and the traffic consumption condition of the target advertisement, so that the automatic dynamic bidding adjustment of the target advertisement is realized. Specifically, first, the target associated resource data of the target advertisement is obtained; the predicted associated resource data of the target advertisement is determined based on the target associated resource data and the predetermined resource constraint data, so that the resource size to be associated with the target advertisement is predicted according to the input resource condition of the target advertisement and the resource constraint set by the advertiser for the target advertisement. Further, the initial bidding price of the target advertisement is determined based on the predicted associated resource data, so that the initial bidding price of the target advertisement can be reasonably set, and the influence of the initially high or low bidding price on the conversion effect is reduced. Then, the bidding exploration is performed based on the initial bidding price, and when it is determined that the actual consumed resource data of the target advertisement meets the predetermined condition, the expected consumed resource data of the target advertisement is determined based on the traffic consumption data and the predetermined resource constraint data, the expected consumed resource of the target advertisement is associated with the actual traffic consumption, which is beneficial to make the expected consumed resource of the target advertisement set as close as possible to the actual traffic consumption, so as to greatly improve the conversion amount of the target advertisement. Finally, the initial bidding price is updated based on the comparison between the actual consumed resource data and the expected consumed resource data, and the exploration condition is set, and the bidding adjustment of the target advertisement is continuously performed according to the difference between the actual resource consumption and the expected resource consumption until the target advertisement meets the predetermined exploration condition, the target bidding price of the target advertisement is determined based on the updated initial bidding price. This way combines the input resource condition of the target advertisement, the pre-set daily resource constraint and the traffic consumption condition to realize the bidding prediction, and dynamically adjusts the bidding according to the difference between the expected consumption and the actual consumption, so as to improve the bidding accuracy, maximize the conversion value of the target advertisement and improve the delivery effect of the target advertisement.

[0073] Other features and advantages of the present disclosure will be set forth in the following description, and in part will become apparent from the description, or can be learned by practice of the present disclosure. The objects and other advantages of the present disclosure can be achieved and obtained by the structures particularly pointed out in the specification, claims and drawings. BRIEF DESCRIPTION OF DRAWINGS

[0074] The accompanying drawings are intended to provide a further understanding of the technical solutions of the present disclosure and constitute a part of the specification, and are used together with the embodiments of the present disclosure to explain the technical solutions of the present disclosure, and do not constitute a limitation on the technical solutions of the present disclosure.

[0075] Figure 1 is a system architecture diagram of an advertisement processing method applied system according to the embodiments of the present disclosure;

[0076] Figures 2A-2EA schematic diagram showing an application of the advertisement processing method according to an embodiment of the present disclosure in an advertisement bidding prediction scenario;

[0077] Figure 3 A flowchart of an advertisement processing method according to an embodiment of the present disclosure;

[0078] Figure 4 A flowchart of determining target associated resource data according to an embodiment of the present disclosure;

[0079] Figure 5 A schematic diagram of an implementation process of determining target associated resource data according to an embodiment of the present disclosure;

[0080] Figure 6 A flowchart of determining predicted associated resource data according to another embodiment of the present disclosure;

[0081] Figure 7 A flowchart of determining an initial bid according to an embodiment of the present disclosure;

[0082] Figures 8A-8C A flowchart of determining an exploration factor of a target advertisement according to an embodiment of the present disclosure;

[0083] Figure 9 A flowchart of judging whether actual consumed resource data meets a predetermined condition according to an embodiment of the present disclosure;

[0084] Figure 10 A flowchart of determining a predetermined bid adjustment factor according to an embodiment of the present disclosure;

[0085] Figure 11 A structural schematic diagram of a comparison table of a weighted sum of a first score and a second score and a predetermined bid adjustment factor according to an embodiment of the present disclosure;

[0086] Figure 12 A flowchart of determining expected consumed resource data of a target advertisement according to an embodiment of the present disclosure;

[0087] Figure 13 A flowchart of fitting processing of traffic consumption data according to an embodiment of the present disclosure;

[0088] Figure 14 A flowchart of determining predicted traffic consumption data according to an embodiment of the present disclosure;

[0089] Figure 15 A flowchart of updating an initial bid according to another embodiment of the present disclosure;

[0090] Figure 16 A flowchart of determining a bid adjustment parameter according to an embodiment of the present disclosure;

[0091] Figure 17 is a flow chart of determining a target bid of a target advertisement based on an updated initial bid according to one embodiment of the present disclosure;

[0092] Figure 18 is a detailed schematic diagram of an advertisement processing method according to one embodiment of the present disclosure;

[0093] Figure 19 is a module diagram of an advertisement processing apparatus according to one embodiment of the present disclosure;

[0094] Figure 20 is a terminal structure diagram of an advertisement processing method according to one embodiment of the present disclosure;

[0095] Figure 21 is a server structure diagram of an advertisement processing method according to one embodiment of the present disclosure. DETAILED DESCRIPTION

[0096] In order to make the objects, technical solutions and advantages of the present disclosure clearer, the present disclosure will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present disclosure and not intended to limit the present disclosure.

[0097] Before the present disclosure is further described, the terms and nouns involved in the embodiments of the present disclosure are explained, and the terms and nouns involved in the embodiments of the present disclosure are applicable to the following explanations:

[0098] Artificial Intelligence (AI) is the theory, method, technology and application system of simulating, extending and expanding human intelligence by using digital computers or machines controlled by digital computers to perceive the environment, acquire knowledge and use knowledge to obtain the best results. In other words, artificial intelligence is a comprehensive technology of computer science, which attempts to understand the essence of intelligence and produce a new intelligent machine that can react in a similar way to human intelligence. Artificial intelligence is to study the design principles and implementation methods of various intelligent machines, so that machines have the functions of perception, reasoning and decision-making. Artificial intelligence technology is a comprehensive discipline, involving a wide range of fields, both hardware and software technologies. Artificial intelligence basic technologies generally include sensors, special artificial intelligence chips, cloud computing, distributed storage, big data processing technology, pre-training model technology, operation / interaction system, mechatronics, etc. Among them, the pre-training model is also called large model, basic model, which can be widely applied to downstream tasks in various directions of artificial intelligence after fine-tuning. Artificial intelligence software technology mainly includes computer vision technology, speech processing technology, natural language processing technology and machine learning / deep learning, etc. With the research and progress of artificial intelligence technology, artificial intelligence technology is researched and applied in many fields, such as common smart home, smart wearable devices, virtual assistants, smart speakers, smart marketing, unmanned vehicles, autonomous vehicles, drones, robots, smart medical care, smart customer service, etc. It is believed that with the development of technology, artificial intelligence technology will be applied in more and more fields and play an increasingly important value.

[0099] Online advertising: also known as web advertising, web marketing, digital advertising or internet advertising, is a form of marketing and advertising that uses the Internet to promote products and services to audiences and platform objects, which can be run by an automated software system across multiple websites, media services and platforms.

[0100] Auction advertising is a new form of Internet advertising in which advertisers compete for advertising exposure by bidding for advertising opportunities. For advertisers, each advertising exposure opportunity can lead to subsequent conversion, and advertisers can obtain advertising exposure by bidding higher resource interactions, thereby bringing more advertising clicks, App activations, user order quantities and more transaction opportunities. Therefore, advertisers want to improve the effectiveness of their ad placements, and need to comprehensively evaluate and analyze different ads and users after bidding to find the most appropriate resource interactions to maximize the conversion of target ads.

[0101] System architecture and scenario description to which the embodiments of the present disclosure are applied

[0102] Figure 1Fig. 1 is a system architecture diagram to which an advertisement processing method according to an embodiment of the present disclosure is applied. It includes object terminals 140, the Internet 130, a gateway 120, an advertisement processing server 110, a traffic database 150, and an advertisement delivery database 160, etc.

[0103] The object terminals 140 include desktop computers, laptop computers, PDAs (personal digital assistants), mobile phones, car terminals, home theater terminals, special-purpose terminals, etc. in various forms. In addition, it can be a single device or a collection of multiple devices. The object terminals 140 can communicate with the Internet 130 in a wired or wireless manner to exchange data. The object terminals 140 contain an advertisement processing platform for an advertiser to set information such as resource constraint quantity of an advertisement bid for a target advertisement, and also for showing the advertiser the dynamic adjustment of the bid for the target advertisement in the advertisement bid.

[0104] The advertisement processing server 110 refers to a computer system capable of providing certain services to the object terminals 140. Compared with the ordinary object terminals 140, the advertisement processing server 110 requires higher stability, security, performance, etc. The advertisement processing server 110 can be a high-performance computer in a network platform, a cluster of multiple high-performance computers, a portion (e.g. a virtual machine) of a high-performance computer, a combination of portions (e.g. virtual machines) of multiple high-performance computers, etc. The advertisement processing server 110 contains various types of services, and the implementation of each service of the advertisement processing server 110 is often associated with some intermediate databases or storage media, etc. The advertisement processing server 110 is used to receive the daily resource constraint information set by the advertiser for the target advertisement in the advertisement bid, and based on the daily resource constraint information set by the advertiser, in combination with the traffic consumption distribution data in the traffic database 150 and the delivery data of the target advertisement in the advertisement delivery database 160, to dynamically adjust the bid for the target advertisement, so as to maximize the conversion quantity of the target advertisement when the daily resource constraint quantity is consumed.

[0105] The gateway 120 is also known as an inter-network connector or a protocol converter. The gateway realizes network interconnection at the transport layer and is a computer system or device that acts as a conversion role. In the case of two systems using different communication protocols, data formats or languages, or even having completely different architectures, the gateway is a translator. At the same time, the gateway can also provide filtering and security functions. The messages sent by the object terminals 140 to the advertisement processing server 110 are sent to the corresponding server through the gateway 120. The messages sent by the advertisement processing server 110 to the object terminals 140 are also sent to the corresponding object terminals 140 through the gateway 120.

[0106] The embodiments of the present disclosure can be applied in various scenarios, for example Figures 2A-2EThe advertisement bidding prediction scenario is shown.

[0107] As Figure 2A shown, when the advertiser clicks the "Confirm" button, a prompt field "The predicted delivery resource amount of the advertisement with the advertisement ID "6659205444" is set to 800 according to the historical delivery data of the advertisement, the initial bid of the advertisement is 400, please confirm" is displayed in the content page of the advertisement processing process, to show the predicted delivery resource amount and the determined initial bid to the advertiser. At this time, the advertiser clicks the "Confirm" button.

[0108] As Figure 2B shown, when the advertiser clicks the "Confirm" button, a prompt field "The predicted delivery resource amount of the advertisement with the advertisement ID "6659205444" is set to 800 according to the historical delivery data of the advertisement, the initial bid of the advertisement is 400, please confirm" is displayed in the content page of the advertisement processing process, to show the predicted delivery resource amount and the determined initial bid to the advertiser. At this time, the advertiser clicks the "Confirm" button.

[0109] As Figure 2C shown, when the advertiser clicks the "Confirm" button, a prompt field "The predicted delivery resource amount of the advertisement with the advertisement ID "6659205444" is set to 800 according to the historical delivery data of the advertisement, the initial bid of the advertisement is 400, please confirm" is displayed in the content page of the advertisement processing process, to show the predicted delivery resource amount and the determined initial bid to the advertiser. At this time, the advertiser clicks the "Confirm" button.

[0110] As Figure 2DAs shown, after automatically allocating the daily resource constraints for the advertisement, a prompt field will be displayed on the advertisement processing flow content page: "The daily resource constraints for the advertisement have been allocated. According to the allocation, the actual resource consumption of the target advertisement at the current moment is less than the allocated resource amount. The target advertisement's resource consumption rate is too slow, and the bid for the target advertisement will be increased to 500." This is to dynamically adjust the advertisement's bid based on the difference between the actual resource consumption and the allocated resource amount for the advertisement ID "6659205444". When the advertisement's resource consumption rate is too slow, the bid for the advertisement will be automatically increased to increase the advertisement's exposure and conversion effect.

[0111] like Figure 2E As shown, after automatically allocating the daily resource constraints for the advertisement, a prompt field will be displayed on the advertisement processing flow content page: "The daily resource constraints for the advertisement have been allocated. According to the allocation, the actual resource consumption of the target advertisement at the current moment is greater than the allocated resource amount. The target advertisement is consuming resources too quickly, and the bid of the target advertisement will be reduced to 450." This is to dynamically adjust the bid of the advertisement based on the difference between the actual resource consumption and the allocated resource amount of the advertisement ID "6659205444". When the advertisement consumes resources too quickly, the bid of the advertisement will be automatically reduced to increase the conversion effect of the advertisement.

[0112] General description of the embodiments of the present disclosure

[0113] According to one embodiment of this disclosure, an advertising processing method is provided.

[0114] This advertising processing method is generally used in business scenarios where the conversion rate of online advertising is highly critical, such as... Figures 2A-2E The illustration shows an advertising bid prediction scenario. This disclosure provides a scheme to determine the difference between the expected and actual resource consumption of an advertisement based on the resource constraints of the target advertisement and the industry's traffic consumption, and to dynamically adjust the bid according to the difference between the expected and actual resource consumption. This improves the accuracy of advertising bids, thereby enhancing the effectiveness of advertising and maximizing advertising conversion rates.

[0115] It should be noted that the dynamic bid adjustment scheme provided in this disclosure aims to maximize the conversion rate or conversion value of the target ad within the daily resource constraint (Budget) given by the advertiser. This objective can be formally expressed as follows:

[0116] max Value;

[0117] stBudget Constraint;

[0118] wherein, max Value refers to the maximization of conversion value; s.t. Budget Constraint refers to the constraint condition of daily resource constraint amount.

[0119] Further, the above formal expression is disassembled into a way of modeling the participation request, wherein the participation request refers to the request of the advertiser participating in the competition of the advertising opportunity, and the above problem can be expressed as follows:

[0120] max xi ∑ i∈Π x i *v i ;

[0121] s.t.∑ i∈Π x i *wp i ≤ Budget;

[0122] wherein, Π refers to the set of participation requests of the advertiser, wp i is the interactive resource amount provided by the successful participant in the i-th participation in the competition of the advertising opportunity, i.e., the minimum bid that the advertiser can successfully compete in the i-th bid request. x i represents whether the advertisement to be bid by the advertiser in the i-th bid wins, when the advertiser wins, x i is 1, when the advertiser does not win, x i is 0. v i refers to the conversion value of the bid advertisement corresponding to the i-th bid request, wherein the conversion value can be determined according to the estimated conversion rate and the estimated exposure click rate of the bid advertisement, i.e., v i＝ pCTR i *pCVR i , pCTR i is the estimated exposure click rate of the bid advertisement corresponding to the i-th bid request, and pCVR i is the estimated conversion rate of the bid advertisement corresponding to the i-th bid request.

[0123] The above s.t.∑ i∈Π x i *wp i ≤ Budget refers to setting the constraint condition as the total of the interactive resource amount provided by all successful (winning) bid requests proposed by the advertiser in a natural day is not greater than the daily resource constraint amount set by the advertiser in the natural day. max xi ∑ i∈Π x i *v iThe objective is to maximize the total conversion value of all successful bids submitted by the advertiser within a given calendar day, under certain constraints.

[0124] Therefore, whether an advertiser can win the competition depends primarily on whether their bid exceeds the amount of interactive resources offered by the winning bidder in this advertising opportunity competition. i Based on this, x i It can be represented as an indicator function: x i =I[pCTR i *pCVR i *bid i wp i ]; where, bid i This refers to the amount of interactive resources (ad bid) offered by the advertiser in the i-th bidding request. According to the GSP's binary pricing principle, this is the amount of interactive resources (wp) offered by the winning bidder in the competition for advertising opportunities. i This also represents the resources that the competitors need to provide to the advertising platform, i.e., the resources they need to consume. Therefore, under the premise that the daily resource consumption meets the daily resource constraint Budget for bidding ads, maximizing the cumulative conversion value (Value) of bidding ads throughout the day can be transformed into finding the optimal bid for bidding ads. i .

[0125] like Figure 3 As shown, an advertising processing method according to an embodiment of this disclosure may include:

[0126] Step 310: Obtain the target-related resource data for the target advertisement;

[0127] Step 320: Based on the target associated resource data and the target ad's pre-defined resource constraint data, determine the target ad's predicted associated resource data, and determine the target ad's initial bid based on the predicted associated resource data;

[0128] Step 330: Explore bids based on the initial bid and determine the actual resource consumption data of the target ad;

[0129] Step 340: If it is determined that the actual resource consumption data meets the predetermined conditions, then based on the pre-acquired traffic consumption data and the predetermined resource constraint data, determine the expected resource consumption data of the target advertisement.

[0130] Step 350, updating the initial bid based on the comparison between the actual consumption resource data and the expected consumption resource data, and returning to the step of determining the expected consumption resource data of the target advertisement based on the preset traffic consumption data and the predetermined resource constraint data until the target advertisement reaches the predetermined exploration condition, determining the target bid of the target advertisement based on the updated initial bid.

[0131] Steps 310-350 are described in detail below.

[0132] In step 310, target associated resource data of the target advertisement is obtained.

[0133] The target advertisement refers to a bid advertisement to be put by the advertiser using the bidding mechanism.

[0134] The target associated resource data is used to indicate the input resource amount of the target advertisement associated with the target dimension.

[0135] The target dimension is used to indicate an item dimension associated with the target advertisement. In the embodiment of the present disclosure, the target dimension can be determined according to the category level of the item pointed to by the target advertisement, for example, the first-level category, the second-level category, the third-level category, the smallest inventory unit (SKU) of the item, and the standardized product unit (SPU) of the item, etc. can be used as the target dimension.

[0136] The input resource amount is used to indicate the size of the input cost of the target advertisement associated with the target dimension.

[0137] In the advertising scenario, the budget of the advertisement is determined based on the input cost of the advertisement and the conversion amount of the advertisement.

[0138] In the implementation of this embodiment, under the condition of authorized permission, all input resource data generated by the target advertisement in the fixed time interval under different optimization targets in the target dimension is obtained first. Then, the input resource mean of the target advertisement under the same optimization target is calculated according to the obtained input resource data. Finally, the input resource mean of the target advertisement in the target dimension is determined as the target associated resource data of the target advertisement.

[0139] In step 320, the predicted associated resource data of the target advertisement is determined based on the target associated resource data and the predetermined resource constraint data of the target advertisement, and the initial bid of the target advertisement is determined based on the predicted associated resource data.

[0140] The predetermined resource constraint data is used to indicate the daily budget of the target advertisement determined by the advertiser in advance.

[0141] The predicted associated resource data is used to indicate the predicted cost of the target advertisement. The predicted associated resource data is used to affect the initial bid of the target advertisement and the upper limit of the actual bid.

[0142] Initial bid is used to indicate the resource interaction amount that an advertiser initially gives when bidding for an advertising opportunity to obtain advertising exposure.

[0143] For brevity, the specific process of determining the predicted associated resource data of the target advertisement based on the target associated resource data and the predetermined resource constraint data of the target advertisement, and the specific process of determining the initial bid of the target advertisement based on the predicted associated resource data will be described in detail below.

[0144] In step 330, bid exploration is performed based on the initial bid, and actual consumed resource data of the target advertisement is determined.

[0145] Bid exploration refers to the process of continuously adjusting the initial bid and determining whether the adjusted initial bid obtains advertising exposure.

[0146] Actual consumed resource data is used to indicate the actual consumption of budget resources by the target advertisement when it is launched.

[0147] In the implementation of this embodiment, first, the cumulative consumed resource amount of the target advertisement in the current natural day is obtained per minute starting from the initial bid of the target advertisement. If the actual cumulative consumed resource amount of the target advertisement under the influence of the initial bid is obtained at the start of the current natural day, it is taken as the actual consumed resource data. Further, if the actual consumed resource data of the target advertisement is zero, it indicates that the target advertisement has not produced exposure, and the initial bid is increased, and the actual cumulative consumed resource amount of the target advertisement in the next minute is obtained, and so on. The exposure of the target advertisement is continuously explored, and the initial bid is adjusted, and the cumulative consumed resource amount of the target advertisement at each minute point is obtained. The actual consumed resource data of the target advertisement can be extracted from the log information of the advertising system under authorized permission.

[0148] In step 340, if it is determined that the actual consumed resource data meets a predetermined condition, the expected consumed resource data of the target advertisement is determined based on the previously obtained traffic consumption data and the predetermined resource constraint data.

[0149] The predetermined condition is used to measure whether the target advertisement has budget consumption when it is launched under the current initial bid.

[0150] Traffic consumption data is used to indicate the proportion of the cumulative traffic consumption at each time point in a natural day to the total traffic consumption of the natural day.

[0151] Expected consumed resource data is used to indicate the consumption of budget resources by the target advertisement when it is launched as expected by the advertiser.

[0152] For brevity, the specific process of determining the expected consumption resource data of the target advertisement based on the pre-acquired traffic consumption data and the predetermined resource constraint data will be described in detail below.

[0153] In step 350, the initial bid is updated based on the comparison of the actual consumption resource data and the expected consumption resource data, and returns to the step of determining the expected consumption resource data of the target advertisement based on the pre-acquired traffic consumption data and the predetermined resource constraint data until the target advertisement reaches the predetermined exploration condition, and the target bid of the target advertisement is determined based on the updated initial bid.

[0154] The predetermined exploration condition is used to measure the degree of dynamic adjustment of the bid for the target advertisement.

[0155] The target bid is used to indicate the final bid of the target advertisement when bidding for the advertising opportunity to obtain advertising exposure. The target bid is generally the bid that can maximize the conversion amount of the target advertisement in the case of successful competition of the advertising opportunity.

[0156] For brevity, the specific process of updating the initial bid based on the comparison of the actual consumption resource data and the expected consumption resource data, and the specific process of determining the target bid of the target advertisement based on the updated initial bid will be described in detail below.

[0157] Through steps 310-350, the embodiments of the present disclosure consider dynamically adjusting the bid of the target advertisement according to the input resource situation of the target advertisement and the traffic consumption situation, which can realize automatic dynamic bid adjustment of the target advertisement, in view of the problem that in the related art, manual setting of the bid of the bid advertisement by the advertiser in the advertisement opportunity bidding process leads to inaccurate bidding. Specifically, first, target associated resource data of the target advertisement is obtained; based on the target associated resource data and predetermined resource constraint data, predicted associated resource data of the target advertisement is determined, which realizes joint prediction of the resource size to be associated with the target advertisement according to the input resource situation of the target advertisement and the resource constraint set by the advertiser for the target advertisement. Further, the initial bid of the target advertisement is determined based on the predicted associated resource data, which can reasonably set the initial bid of the target advertisement and reduce the influence of the initial bid that is too high or too low on the conversion effect. Then, bid exploration is performed based on the initial bid, and when it is determined that the actual consumed resource data of the target advertisement meets a predetermined condition, the expected consumed resource data of the target advertisement is determined based on the traffic consumption data and the predetermined resource constraint data, which associates the expected consumed resource of the target advertisement with the actual traffic consumption, which is conducive to making the expected consumed resource of the target advertisement set as close as possible to the actual traffic consumption, so as to greatly improve the conversion amount of the target advertisement. Finally, the initial bid is updated based on the comparison of the actual consumed resource data and the expected consumed resource data, and an exploration condition is set, and the target advertisement is continuously bid adjusted according to the difference between the actual resource consumption and the expected resource consumption until the target advertisement meets the predetermined exploration condition, and the target bid of the target advertisement is determined based on the updated initial bid. This way combines the input resource situation of the target advertisement, the pre-set daily resource constraint, and the traffic consumption situation to realize bid prediction, and dynamically adjusts the bid according to the difference between the expected consumption and the actual consumption, thereby improving the accuracy of the bid, which helps to maximize the conversion value of the target advertisement and improve the delivery effect of the target advertisement.

[0158] The above is a general description of steps 310-350. Since step 330 has been described in detail above, the specific implementation of steps 310, 320, 340, and 350 will be described in detail below.

[0159] Detailed description of step 310

[0160] In step 310, target associated resource data of the target advertisement is obtained, wherein the target associated resource data is used to indicate the input resource amount associated with the target advertisement in the target dimension.

[0161] In the implementation of this embodiment, the input resource of the target advertisement in each target dimension is often recorded in the advertisement system log. Based on this, the input resource data of the target advertisement in each target dimension can be obtained from the advertisement system log under the condition of authorized permission.

[0162] Since the conversion cost of different optimization targets is often different, for example, the optimization cost of the shallow layer (the advertisement guiding each user to perform the activation operation, etc.) is generally lower than that of the deep layer (the advertisement guiding each user to make a payment, etc.). In addition, there is a certain difference in the conversion cost of different types of advertisements. For example, the conversion cost of the advertisement of the game industry is different from that of the advertisement of the clothing industry. Based on this, the disclosure embodiment improves a scheme of determining the target associated resource data in the order from bottom to top based on the hierarchical relationship of each dimension and the exposure of the target advertisement in each dimension, which can improve the accuracy and reliability of the target associated resource data.

[0163] It should be noted that the target dimension of the target advertisement in the embodiment of the disclosure can be any one of the second-level category of the item, the third-level category of the item, the smallest inventory unit (SKU) of the item, and the standardized product unit (SPU) of the item. Among them, the level of the second-level category of the item is higher than the level of the third-level category of the item, the level of the third-level category of the item is higher than the level of the standardized product unit (SPU) of the item, and the level of the standardized product unit (SPU) of the item is higher than the level of the smallest inventory unit (SKU) of the item.

[0164] For example, when the target advertisement is for the game industry, the game is the second-level category; the game can also be divided into 2D games and 3D games, and the 2D games and 3D games are the third-level category. For 2D games and 3D games, the games subdivided in different times and different batches are released, so for 2D games and 3D games, the standardized product unit (SPU) of the item can be obtained according to the release batch. Further, for multiple games in the same batch, the dimension of each game is the smallest inventory unit (SKU) of the item.

[0165] In one specific example, for a 2D game (game 1) released in the third batch, first obtain the target exposure data of the game 1, which can be the number of objects that activate the game 1.

[0166] Please refer to Figure 4 In one embodiment, step 310 specifically includes but is not limited to steps 410-440:

[0167] Step 410, obtaining target exposure data of the target advertisement in the target dimension;

[0168] If it is determined that the target exposure data does not reach the predetermined exposure threshold, obtaining reference exposure data of the target advertisement in a higher level dimension of the target dimension in step 420.

[0169] Replacing the target dimension with the higher level dimension of the target dimension and replacing the target exposure data with the reference exposure data in step 430, and returning to the step of obtaining the reference exposure data of the target advertisement in the higher level dimension of the target dimension if it is determined that the target exposure data does not reach the predetermined exposure threshold, until the target exposure data reaches the predetermined exposure threshold, obtaining the input resource average value associated with the target dimension;

[0170] Determining the target associated resource data based on the input resource average value in step 440.

[0171] The steps 410-440 are described in detail as follows.

[0172] In step 410, obtaining the target exposure data of the target advertisement under the target dimension.

[0173] The target exposure data is used to indicate the advertisement conversion of the target advertisement under the target dimension.

[0174] In the implementation of this embodiment, the target exposure data of the target advertisement under the target dimension is recorded in the background log of the advertisement system. Based on this, the target exposure data of the target advertisement under the target dimension is obtained from the background log of the advertisement system under the authorization permission.

[0175] In step 420, if it is determined that the target exposure data does not reach the predetermined exposure threshold, obtaining the reference exposure data of the target advertisement in a higher level dimension of the target dimension.

[0176] The predetermined exposure threshold is used to measure whether the target exposure data is reliable.

[0177] The reference exposure data of the higher level dimension is used to indicate the advertisement conversion of the target advertisement under the higher level dimension of the current target dimension.

[0178] In the implementation of this embodiment, first, the target exposure data is compared with the predetermined exposure threshold. If it is determined that the target exposure data does not reach the predetermined exposure threshold, it indicates that the exposure data of the target advertisement in the current target dimension is not reliable and cannot be used as a reference. Based on this, first, according to the hierarchical relationship of each dimension, the higher level dimension of the current target dimension is determined. Then, the exposure data of the target advertisement in the higher level dimension is obtained as the reference exposure data under the authorization permission.

[0179] In step 430, the target dimension is replaced by the upper level dimension of the target dimension, and the target exposure data is replaced by the reference exposure data, and the step of obtaining the reference exposure data of the target advertisement in the upper level dimension of the target dimension is returned if it is determined that the target exposure data does not reach the predetermined exposure threshold, until the target exposure data reaches the predetermined exposure threshold, the input resource mean value associated with the target dimension of the target advertisement is obtained.

[0180] The input resource mean value is used to indicate the average cost of the target advertisement in a target dimension.

[0181] In the embodiment, first, the target dimension is replaced by the upper level dimension of the target dimension, and the target exposure data is replaced by the reference exposure data. Then, the confidence of the target advertisement in the new target dimension is judged. If it is determined that the target exposure data does not reach the predetermined exposure threshold, it indicates that the current target dimension is not confident, and the process is repeated until the target exposure data of the current target dimension reaches the predetermined exposure threshold, indicating that the target advertisement in the current target dimension is confident and can be used as a reference. Based on this, first, the generated cost data of the target advertisement in the same optimization target in the current target dimension is obtained. Then, the total number of conversions of the target advertisement in the current target dimension is obtained. Finally, the quotient of the generated cost data and the total number of conversions is taken as the input resource mean value associated with the target dimension of the target advertisement.

[0182] For example, for a target advertisement in the game industry, when the optimization target is to guide each object to activate game 1, the cumulative input resource generated in 7 days is counted when the target advertisement is counted as the optimization target to guide each object to activate the third batch of a 2D game (game 1), and the total number of objects activating game 1 in 7 days is counted. The quotient of the cumulative input resource generated in 7 days and the total number of objects activating game 1 in 7 days is taken as the input resource mean value.

[0183] In step 440, the target associated resource data is determined based on the input resource mean value.

[0184] In the embodiment, in order to improve the efficiency of the bid prediction, the input resource mean value can be directly determined as the target associated resource data. The target associated resource data is the estimated cost benchmark data of the target advertisement.

[0185] As Figure 5As shown, the hierarchical relationship of each level of the embodiment of the present disclosure is: the first level dimension of the product to which the advertisement belongs (the second level category of the product) is higher than the second level dimension of the product to which the advertisement belongs (the third level category of the product); the second level dimension of the product to which the advertisement belongs (the third level category of the product) is higher than the third level dimension of the product to which the advertisement belongs (the SPU of the product); and the third level dimension of the product to which the advertisement belongs (the SPU of the product) is higher than the fourth level dimension of the product to which the advertisement belongs (the SKU of the product). Based on this, the fallback order when obtaining the target associated resource data is from the SKU of the product to the SPU of the product, from the SPU of the product to the third level category of the product, and finally from the third level category of the product to the second level category of the product, which is a fallback mode from fine granularity to coarse granularity.

[0186] For example, the target advertisement is for an electronic product, the target dimension is first determined as the smallest inventory unit (SKU) of the product, and the target exposure data can be the cumulative conversion number of the electronic product within 7 days. When the cumulative conversion number reaches 10, it is considered that the target exposure data is confident, and the average delivery cost of the cumulative conversion number of electronic products is taken as the target associated resource data. When the cumulative conversion number does not reach 10, it is considered that the target exposure data is not confident, and the target dimension is modified to the standardized product unit (SPU) of the product, and the cumulative conversion number of all electronic products of the product batch to which the electronic product belongs within 7 days is obtained as new target exposure data. When the cumulative conversion number reaches 10, it is considered that the target exposure data is confident, and the average delivery cost of the cumulative conversion number of electronic products converted in the same product batch is taken as the target associated resource data. When the cumulative conversion number does not reach 10, it is considered that the target exposure data is not confident, and the target dimension is continued to be rolled back to the upper dimension until the second level category of the product is taken as the target dimension.

[0187] The advantage of this embodiment is that the hierarchical relationship based on each dimension is considered, and the target associated resource data is determined in combination with the exposure of the target advertisement in each dimension. Starting from the delivery data and exposure data of the target advertisement in the lowest level dimension, when the exposure data is not confident, the exposure of the target advertisement is judged in the order from bottom to top, and the average of the input resources in the dimension whose exposure is confident is taken as the target associated resource data, which realizes the target associated resource data acquisition from fine granularity to coarse granularity, and can improve the accuracy and reliability of the target associated resource data.

[0188] Detailed description of step 320

[0189] In step 320, the predicted associated resource data of the target advertisement is determined based on the target associated resource data and the predetermined resource constraint data of the target advertisement, and the initial bid of the target advertisement is determined based on the predicted associated resource data.

[0190] The size of the predicted input resource for the target advertisement often affects the initial bid and the actual bid upper limit of the target advertisement. When the predicted input resource for the target advertisement is too small, the dynamic adjustment of the bid is limited, which affects the initial amount of the target advertisement. When the predicted input resource for the target advertisement is too large, the dynamic adjustment of the bid is too aggressive, which easily leads to rapid resource consumption and low resource utilization. Based on this, the embodiments of the present disclosure provide a scheme for determining the predicted input resource of the target advertisement based on the budget constraint of the target advertisement and the generated delivery data as a reference, which can improve the rationality and accuracy of the predicted associated resource data.

[0191] Please refer to Figure 6 In one embodiment, the specific process of determining the predicted associated resource data of the target advertisement in step 320 includes but is not limited to the following steps 610-620:

[0192] Step 610, determining the number of associated items of the target advertisement generated in one time day under the target dimension;

[0193] Step 620, determining the predicted associated resource data based on the number of associated items, the target associated resource data, and the predetermined resource constraint data.

[0194] The steps 610-620 are described in detail below.

[0195] In step 610, the number of associated items of the target advertisement generated in one time day under the target dimension is determined.

[0196] One time day refers to one natural day.

[0197] The number of associated items refers to the cumulative conversion number of the items associated with the target advertisement in one natural day.

[0198] For example, for a target advertisement in the game industry, when the optimization target is to guide each object to activate game 1, the number of associated items is the total number of objects that activate game 1 in one natural day.

[0199] In the implementation of this embodiment, under the condition of authorized permission, first, the total number of conversions of the target advertisement in a continuous fixed number of natural days under the target dimension is extracted from the log data of the advertisement system. Then, the total number of conversions is divided by the fixed number to obtain the number of associated items of the target advertisement generated in one time day under the target dimension.

[0200] In step 620, the predicted associated resource data is determined based on the associated item number, the target associated resource data, and the predetermined resource constraint data.

[0201] In the implementation of this embodiment, step 620 specifically includes:

[0202] determining a first product based on the associated item number and the target associated resource data;

[0203] performing logarithmic scaling on a comparison result of the predetermined resource constraint data and the first product to obtain a logarithmic scaling result;

[0204] performing scaling processing on the logarithmic scaling result and 1 and the target associated resource data to obtain the predicted associated resource data.

[0205] Specifically, first, the associated item number and the target associated resource data are multiplied to obtain a first product, where the first product is used to indicate the average resource consumption of the target advertisement under the current target dimension. Then, the predetermined resource constraint data and the first product are divided to obtain a comparison result of the predetermined resource constraint data and the first product, where the comparison result is used to indicate the difference degree of the predetermined resource constraint data and the first product. Further, the comparison result is taken logarithm to realize logarithmic scaling of the comparison result, and a logarithmic scaling result is obtained. Finally, the logarithmic scaling result and the first sum of 1 and the target associated resource data are multiplied to obtain the predicted associated resource data.

[0206] In the embodiments of the present disclosure, the calculation process of the predicted associated resource data can be represented as formula (1):

[0207]

[0208] Wherein, CPA is the predicted associated resource data; historyCPA is the target associated resource data; T is the associated item number; Budget is the predetermined resource constraint data.

[0209] As can be seen, when the goal is to consume all the constraint resources set in the predetermined resource constraint data, if the first product is less than the predetermined resource constraint data, it indicates that the cost of the target advertisement needs to be set to a larger value, so that the predicted associated resource data is greater than the target associated resource data to obtain more conversion volume.

[0210] The advantage of this embodiment is that based on the budget constraint of the target advertisement and the generated delivery data as a reference to determine the predicted input resource of the target advertisement, the rationality and accuracy of the predicted associated resource data can be improved.

[0211] Please refer to Figure 7In one embodiment, the specific process of determining the initial bid of the target advertisement based on the predicted associated resource data comprises but is not limited to steps 710-720:

[0212] Step 710, determining an exploration factor of the target advertisement based on a starting speed of the target advertisement and an associated input resource;

[0213] Step 720, determining the predicted associated resource data based on the associated item number, the target associated resource data, and predetermined resource constraint data.

[0214] The steps 710-720 are described in detail as follows.

[0215] In step 710, an exploration factor of the target advertisement is determined based on a starting speed of the target advertisement and an associated input resource.

[0216] The starting speed is used to indicate the number of conversions accumulated by the target advertisement in a time unit, and is used to measure the conversion efficiency of the target advertisement at the beginning of the delivery.

[0217] The associated input resource is used to indicate the resource consumption of the target advertisement.

[0218] The exploration factor is used to indicate the relationship between the initial bid of the target advertisement and the predicted associated resource data.

[0219] In the implementation of this embodiment, first, a first table indicating the corresponding relationship between each speed interval and score is obtained for the starting speed of the target advertisement. Then, the speed interval in which the starting speed of the target advertisement is located is found in the first table, and the score corresponding to the found speed interval is taken as the first predetermined score of the target advertisement. Similarly, a second table indicating the corresponding relationship between each input resource interval and score is obtained for the associated input resource of the target advertisement. Then, the input resource interval in which the associated input resource of the target advertisement is located is found in the second table, and the score corresponding to the found input resource interval is taken as the second predetermined score of the target advertisement. Finally, the first predetermined score and the second predetermined score are weighted and added according to a preset weight ratio to obtain the total score of the target advertisement, and the exploration factor of the target advertisement is determined according to the total score. The specific process of determining the exploration factor of the target advertisement according to the total score is similar to the process of determining the first predetermined score. For the sake of brevity, no longer described.

[0220] For example, Figure 8AThe table shown is a comparison table of the initial volume rate and the first predetermined score. Specifically, when the initial volume rate is in speed range 1, the first predetermined score is 80; when the initial volume rate is in speed range 2, the first predetermined score is 70; when the initial volume rate is in speed range 3, the first predetermined score is 60; when the initial volume rate is in speed range 4, the first predetermined score is 50; and when the initial volume rate is in speed range 5, the first predetermined score is 60. The initial volume rate decreases sequentially from speed range 1 to speed range 5.

[0221] like Figure 8B The table shown is a comparison table of associated input resources and the second predetermined score. Specifically, when the associated input resources are in resource quantity interval 1, the second predetermined score is 80; when the associated input resources are in resource quantity interval 2, the second predetermined score is 70; when the associated input resources are in resource quantity interval 3, the second predetermined score is 60; when the associated input resources are in resource quantity interval 4, the second predetermined score is 50; and when the associated input resources are in resource quantity interval 5, the second predetermined score is 40. The associated input resources decrease sequentially from resource quantity interval 1 to resource quantity interval 5.

[0222] like Figure 8C The table shown is a comparison table of total score and exploration factor. Specifically, when the total score is 81-100, the exploration factor is 2; when the total score is 61-80, the exploration factor is 3; when the total score is 41-60, the exploration factor is 4; when the total score is 21-40, the exploration factor is 5; and when the total score is 0-20, the exploration factor is 6.

[0223] In step 720, the initial bid is determined based on the inverse of the exploration factor and the product of the predicted associated resource data.

[0224] In this specific implementation, firstly, the reciprocal of the exploration factor is calculated. Then, the reciprocal of the exploration factor is multiplied by the predicted associated resource data to obtain the product, which is used as the initial bid. The initial bid can be expressed as: Where n is the exploration factor, n>1. bid0 is the initial bid of the target ad in a campaign period, which refers to one calendar day. bid0 is the resource interaction volume of the target ad at the beginning of a calendar day.

[0225] For example, when the predicted associated resource data for a target ad is 100 and the exploration factor is 4, the initial bid for the target ad is 25. Within a calendar day, starting with 25 resources, the system explores the resource interaction volume that maximizes the conversion rate of the target ad, assuming the predetermined resource constraints are exhausted. Here, exploration refers to increasing or decreasing the initial bid.

[0226] The advantage of this embodiment is that, in the way of introducing the table lookup method and the fractional weighting, the exploration factor of the target advertisement is determined based on the initial speed of the target advertisement and the associated input resources, and then the initial bid of the target advertisement is set according to the exploration factor and the predicted associated resource data, which can effectively avoid that the initial bid is set too high or too low, and can improve the rationality and accuracy of the exploration starting position of the target advertisement.

[0227] Detailed description of the adjustment of the initial bid based on the bid exploration situation in the embodiments of the present disclosure

[0228] In actual application, if the initial resource interaction (initial bid) of the target advertisement is too small, it will lead to that the target advertisement is not enough to produce exposure, and often cannot achieve the maximization of the conversion amount of the target advertisement under the condition of consuming the specified resources in the resource constraint data. Based on this, the embodiment of the present disclosure provides a scheme for adjusting the initial bid based on the exposure of the target advertisement, which can improve the adjustment flexibility of the initial bid.

[0229] Please refer to Figure 9 In one embodiment, after determining the actual consumed resource data of the target advertisement, the advertisement processing method further includes but is not limited to the following steps 910-930:

[0230] Step 910, determining whether the actual consumed resource data is non-zero data;

[0231] Step 920, if it is determined that the actual consumed resource data is non-zero data, determining whether the actual consumed resource data meets a predetermined condition;

[0232] Step 930, if it is determined that the actual consumed resource data is not non-zero data, performing bid adjustment on the initial bid based on a predetermined bid adjustment factor until the actual consumed resource data is non-zero data, and replacing the initial bid with the adjusted initial bid.

[0233] The steps 910-930 are described in detail below.

[0234] In step 910, it is determined whether the actual consumed resource data is non-zero data.

[0235] The non-zero data is used to indicate that the target advertisement has consumed resources in the delivery, and the target advertisement has produced exposure.

[0236] In the implementation of this embodiment, when the target advertisement produces exposure, there will be consumption of resources, at this time the actual consumption data should be a non-zero value, and when the target advertisement does not produce exposure in the delivery, there will be no consumption of resources, and the actual consumption data is zero. Based on this, the actual consumption data is numerically tested to determine whether the actual consumed resource data is non-zero data,

[0237] In step 920, if it is determined that the actual consumed resource data is non-zero data, it is determined that the actual consumed resource data satisfies the predetermined condition.

[0238] In the embodiment, if it is determined that the actual consumed resource data is non-zero data, it indicates that the current initial bid can make the target advertisement produce exposure, the target advertisement appears to consume resources, and the initial bid does not exist the problem of being too small, so it is determined that the actual consumed resource data satisfies the predetermined condition.

[0239] In step 930, if it is determined that the actual consumed resource data is not non-zero data, the initial bid is bid adjusted based on the predetermined bid adjustment factor until the actual consumed resource data is non-zero data, and the initial bid is replaced by the adjusted initial bid.

[0240] The predetermined bid adjustment factor is used to indicate the adjustment amplitude of the initial bid.

[0241] The adjusted initial bid is used to indicate the initial bid that can make the target advertisement produce exposure in delivery.

[0242] In the embodiment, if it is determined that the actual consumed resource data is not non-zero data, it indicates that the current initial bid is too small and is not sufficient to make the target advertisement produce exposure, and the initial bid needs to be adjusted to a larger value. Based on this, at each minute point of a natural day, the initial bid is bid adjusted based on the predetermined bid adjustment factor to obtain the adjusted initial bid of the current minute point, and it is continuously determined whether the target advertisement produces exposure and appears to consume resources under the adjusted initial bid, and this process is repeatedly performed until at a certain minute point, the target advertisement produces exposure and appears to consume resources, so that the actual consumed resource data is non-zero data, and the initial bid is replaced by the adjusted initial bid of this minute point.

[0243] The initial bid is bid adjusted based on the predetermined bid adjustment factor, including:

[0244] Determining a second sum of the predetermined bid adjustment factor and 1;

[0245] Determining the adjusted initial bid based on the product of the second sum and the initial bid.

[0246] In the embodiment, first, the second sum of the predetermined bid adjustment factor and 1 is obtained, and the second sum is used to indicate the size relationship between the adjusted initial bid of the current minute point and the initial bid of the previous minute point of the current minute point. Then, the second sum is multiplied by the initial bid, and the product of the second sum and the initial bid is taken as the adjusted initial bid of the current minute point.

[0247] The adjusted initial bid for each minute point can be represented as shown in formula (2):

[0248] bid t = (1+μ)*bid t-1 Formula (2);

[0249] Among them, bid t This refers to the initial bid after adjustment at the current minute. t-1 This refers to the initial bid adjusted from the previous minute. μ refers to the pre-set bid adjustment factor, with a value range of [0.01, 0.05].

[0250] For example, at 7:00 AM on a calendar day, the current minute represents the accumulated minutes from midnight to 7:00 AM on that day. The current minute is t = 420. The previous minute was 6:59 AM, and the minute before that was t-1 = 419. Therefore, the adjusted initial bid at 7:00 AM is bid. 420 bid 420 = (1+μ)*bid 419 .

[0251] The advantage of this embodiment is that the initial bid is adjusted based on the exposure of the target ad. The initial bid is adjusted every minute of a natural day. When the initial bid is too low to generate exposure for the target ad, the initial bid is continuously increased to increase the value of the initial bid, accelerate the consumption of resources by the target ad, and enable the target ad to generate exposure quickly. This improves the flexibility of adjusting the initial bid.

[0252] Please refer to Figure 10 In one embodiment, the predetermined bid adjustment factor is determined in the following manner:

[0253] Step 1010: Determine the difficulty level of the target ad and the placement scenario;

[0254] Step 1020: Determine the first score based on the difficulty level of the competition;

[0255] Step 1030: Determine the second score based on the delivery scenario;

[0256] Step 1040: Determine the pre-bid adjustment factor based on the first score and the second score.

[0257] Steps 1010-1040 are described in detail below.

[0258] In step 1010, the difficulty level of winning the target advertisement and the delivery scenario are determined.

[0259] The winning difficulty level is used to indicate the difficulty of success of the target advertisement in the advertisement opportunity bidding. When the number of participants in the advertisement opportunity bidding is larger, the winning difficulty level is higher; when the number of participants in the advertisement opportunity bidding is smaller, the winning difficulty level is lower.

[0260] The placement scenario is used to indicate the specific application scenario to which the target advertisement is to be placed.

[0261] In the implementation of this embodiment, first, the total number of participants in the advertisement opportunity bidding together with the target advertisement to be placed by the advertiser is counted. Then, according to the comparison between the total number and the predetermined number interval corresponding to each winning difficulty level, the winning difficulty level corresponding to the predetermined number interval in which the total number is located is taken as the winning difficulty level of the target advertisement. Further, the advertisement basic information of the target advertisement is obtained from the server background under the authorized permission, and the placement scenario of the target advertisement is extracted from the advertisement basic information.

[0262] In step 1020, the first score is determined based on the winning difficulty level.

[0263] In the implementation of this embodiment, the specific process of step 1020 is similar to the process of determining the first predetermined score based on the starting speed in step 710 described above. For the sake of brevity, no longer be described.

[0264] In step 1030, the second score is determined based on the placement scenario.

[0265] In the implementation of this embodiment, the specific process of step 1030 is similar to the process of determining the first predetermined score based on the starting speed in step 710 described above. For the sake of brevity, no longer be described.

[0266] In step 1040, the predetermined bid adjustment factor is determined based on the first score and the second score.

[0267] In the implementation of this embodiment, the specific process of step 1040 is similar to the process of determining the exploration factor based on the first predetermined score and the second predetermined score in step 710 described above. For the sake of brevity, no longer be described.

[0268] The value range of the predetermined bid adjustment factor can be set between 0.01 and 0.05.

[0269] As Figure 11The table shows the weighted sum of the first score and the second score and the predetermined bid adjustment factor. Specifically, when the total score is 81-100, the predetermined bid adjustment factor is 0.05; when the total score is 61-80, the predetermined bid adjustment factor is 0.04; when the total score is 41-60, the predetermined bid adjustment factor is 0.03; when the total score is 21-40, the predetermined bid adjustment factor is 0.02; and when the total score is 0-20, the predetermined bid adjustment factor is 0.01.

[0270] The advantage of this embodiment is that, according to the competition difficulty level and the delivery scenario, the scoring mechanism and the weighted calculation method are introduced to determine the predetermined bid adjustment factor, which can set the predetermined bid adjustment factor to a larger value when the competition difficulty level is high and in the new material or new product test scenario, the key position investment scenario and other delivery scenarios that require fast exposure of target advertisements, so as to quickly increase the initial bid of the target advertisement, so that the target advertisement can produce exposure under the initial bid, and the adjustment flexibility and accuracy of the predetermined bid adjustment factor can be improved.

[0271] Detailed description of step 340

[0272] In step 340, if it is determined that the actual consumption resource data meets the predetermined condition, the expected consumption resource data of the target advertisement is determined based on the pre-acquired traffic consumption data and the predetermined resource constraint data.

[0273] Please refer to Figure 12 In one embodiment, step 340 specifically includes but is not limited to the following steps 1210-1250:

[0274] Step 1210, fitting processing is performed on the traffic consumption data to obtain a first curve fitting result for indicating the traffic consumption distribution of the target time day and a second curve fitting result for indicating the traffic consumption distribution of the non-target time day;

[0275] Step 1220, determining the current time day and the cumulative time data of the current time day;

[0276] Step 1230, based on the current time day, determining the target curve fitting result from the first curve fitting result and the second curve fitting result;

[0277] Step 1240, based on the cumulative time data, using the target curve fitting result to determine the predicted traffic consumption data.

[0278] Step 1250, based on the product of the predicted traffic consumption data and the predetermined resource constraint data, determining the expected consumption resource data of the target advertisement.

[0279] The steps 1210-1250 are described in detail below.

[0280] In step 1210, the flow consumption data is fitted to obtain a first curve fitting result for indicating a flow consumption distribution of a target time day, and a second curve fitting result for indicating a flow consumption distribution of a non-target time day.

[0281] The target time day refers to a working day in a natural day.

[0282] The non-target time day refers to a non-working day (i.e. weekend) in a natural day.

[0283] The flow consumption distribution is used to indicate flow consumption in each minute of a natural day.

[0284] The first curve fitting result is used to indicate a mathematical operation form of flow consumption in each minute point in the target time day.

[0285] The second curve fitting result is used to indicate a mathematical operation form of flow consumption in each minute point in the non-target time day.

[0286] For brevity, the specific process of fitting the flow consumption data in the embodiment of the present disclosure will be described in detail hereinafter, and will not be repeated here.

[0287] In step 1210, the current time day and the accumulated time data of the current time day are determined.

[0288] The current time day refers to a natural day in which the ad opportunity bidding of the target advertisement is located.

[0289] The accumulated time data is used to indicate the number of minutes accumulated from zero point to the current time in the current time day. For example, if the current time is 12:00 noon, the accumulated time data is the number of minutes accumulated from zero point to 12:00 noon, and the accumulated time data (current minute point) is t=720.

[0290] In the implementation of the embodiment, first, in the case of authorized permission, the current time day and the number of minutes passed in the current time day are queried in the time dimension by using a calendar or an equal time recording tool to obtain the date indicated by the current time day and the number of minutes passed in the current time day. Then, the number of minutes passed in the current time day is recorded as the accumulated time data of the current time day.

[0291] In step 1230, based on the current time day, the target curve fitting result is determined from the first curve fitting result and the second curve fitting result.

[0292] In the implementation of the embodiment, first, time category judgment is performed on the current time day to determine whether the current time day is a weekday or a weekend. Then, if it is determined that the current time day is a weekday, it is indicated that the current time day is a target time day, and the first curve fitting result is determined as the target curve fitting result. If it is determined that the current time day is a weekend, it is indicated that the current time day is a non-target time day, and the second curve fitting result is determined as the target curve fitting result.

[0293] In step 1240, based on the accumulated time data, the target curve fitting result is used to determine the predicted traffic consumption data.

[0294] In the implementation of the embodiment, since the target curve fitting result can indicate the operation manner of the proportion of the accumulated traffic consumption of each accumulated minute of the current time day to the whole day traffic consumption. Based on this, the accumulated time data is substituted into the target curve fitting result as the independent variable to calculate the proportion of the accumulated traffic consumption of the accumulated time data of the current time day to the whole day traffic consumption, and the calculated proportion is taken as the predicted traffic consumption data.

[0295] In step 1250, based on the product of the predicted traffic consumption data and the predetermined resource constraint data, the expected consumption resource data of the target advertisement is determined.

[0296] In the implementation of the embodiment, for each minute point of a natural day, the expected consumption resource data of the target advertisement at the current minute point is obtained by multiplying the predicted traffic consumption data corresponding to the current minute point (current accumulated time data) and the predetermined resource constraint data.

[0297] Wherein, the expected consumption resource data of the target advertisement at the current minute point t can be expressed as expectCost t = Budget*F(t), expectCost t is the expected consumption resource data of the current minute point t; Budget is the predetermined resource constraint data; F(t) is the predicted traffic consumption data corresponding to the current minute point (current accumulated time data).

[0298] The advantage of the embodiment is that the traffic consumption data is fitted to obtain the traffic curve fitting results corresponding to the weekdays and weekends respectively, the traffic consumption of each minute point in a natural day is determined by using mathematical method, the predetermined resource constraint data of a natural day is divided into resource constraints corresponding to each minute point according to the traffic consumption of each minute point, the effect of resource all-day distribution of the target advertisement is achieved, which is beneficial to making the resource consumption of the target advertisement more close to the traffic consumption curve in reality, so as to maximize the conversion amount of the target advertisement as much as possible.

[0299] Please refer to Figure 13 In one embodiment, step 1210 specifically includes, but is not limited to, steps 1310-1330:

[0300] Step 1310, based on the flow consumption data, determining the first proportion of the cumulative flow consumption at each sampling time point of the target time day in the total flow consumption of the target time day, and determining the second proportion of the cumulative flow consumption at each sampling time point of the non-target time day in the total flow consumption of the non-target time day;

[0301] Step 1320, based on the first proportion of each sampling time point of the target time day, using the least square method to determine the first curve fitting result;

[0302] Step 1330, based on the second proportion of each sampling time point of the non-target time day, using the least square method to determine the second curve fitting result.

[0303] The steps 1310-1330 are described in detail below.

[0304] In step 1310, based on the flow consumption data, the first proportion of the cumulative flow consumption at each sampling time point of the target time day in the total flow consumption of the target time day is determined, and the second proportion of the cumulative flow consumption at each sampling time point of the non-target time day in the total flow consumption of the non-target time day is determined.

[0305] In order to improve the fitting effect, the flow consumption data of the embodiments of the present disclosure is the flow consumption data of each sampling time point of each natural day within N days of the time interval of the current time day, wherein N is an integer greater than 30.

[0306] The sampling time point is used to indicate the minute point of the cumulative flow consumption in each natural day.

[0307] The cumulative flow consumption is used for the total flow consumption from zero to the sampling time point in a natural day.

[0308] The total flow consumption is used for the total flow consumption from the beginning to the end of a natural day.

[0309] The first proportion is used to indicate the real proportion of the flow consumption at each sampling time point of the working day in the total flow consumption of the day.

[0310] The second proportion is used to indicate the real proportion of the flow consumption at each sampling time point of the weekend in the total flow consumption of the day.

[0311] In the implementation of this embodiment, first, considering that the flow consumption is quite different between weekdays and non-weekdays, first, the flow consumption data is divided into flow consumption data of weekdays and flow consumption data of non-weekdays (weekends) according to the time day to which the flow consumption data belongs. Then, for each sampling time point of each target time day, the cumulative flow consumption at the current sampling time point and the total flow consumption of the target time day are extracted from the flow consumption data. Finally, the cumulative flow consumption at the current sampling time point is divided by the total flow consumption of the target time day to obtain the first proportion of the current sampling time point.

[0312] Similarly, the specific process of determining the second proportion in the embodiment of the disclosure is similar to the specific process of determining the first proportion of each sampling time point of the target time day described above. To save space, no longer be described.

[0313] In step 1320, based on the first proportion of each sampling time point of the target time day, the least square method is used to determine the first curve fitting result.

[0314] In the implementation of this embodiment, considering that the flow consumption is quite different in different time periods of a natural day, for example, between zero and six o'clock, the activity frequency of each object is low, and the flow consumption is often small, while between two o'clock in the afternoon and twelve o'clock at night, the activity frequency of each object is higher, and the flow consumption is often larger. Based on this, first, the cumulative minutes of a natural day are divided into multiple intervals, and each interval includes multiple sampling time points. Then, for each interval, the least square method is used to jointly fit the multiple sampling time points and their first proportions to obtain a flow consumption curve equation of the target time day in the interval, taking each sampling time point as the independent variable and the first proportion corresponding to each sampling time point as the dependent variable. Finally, the flow consumption curve equations of the multiple intervals are collectively determined as the first curve fitting result.

[0315] Further, in order to improve the accuracy of the fitting result, a predetermined threshold can also be set for some intervals to compare the output of the flow consumption curve equation, and the larger or smaller one is taken as the final output of the first curve fitting result.

[0316] Wherein, the first curve fitting result can be represented as formula (3):

[0317]

[0318] Wherein, F workday (t) is the predicted proportion of flow consumption at each cumulative minute of the weekdays to the total flow consumption of the day. T is the number of minutes that have passed in the current natural day.

[0319] Specifically, when the accumulated minutes are between (0, 360], the proportion of the accumulated flow consumption in the total flow consumption at the current accumulated minutes is calculated according to the accumulated minutes t by using the operation formula “-8.875*10 -7 t 2 +6.527*10 -4 t”, and then the calculated proportion is compared with a predetermined value (5*10 -5 ), and the larger one is taken as the flow consumption proportion corresponding to the current accumulated minutes output by the first curve fitting result.

[0320] When the accumulated minutes are between (840, 1440], the proportion of the accumulated flow consumption in the total flow consumption at the current accumulated minutes is calculated according to the accumulated minutes t by using the operation formula “3.334*10 -7 t 2 +1.681*10 -4 t+0.076”, and then the calculated proportion is compared with a predetermined value (1.0), and the smaller one is taken as the flow consumption proportion corresponding to the current accumulated minutes output by the first curve fitting result.

[0321] When the accumulated minutes are between (360, 840], the proportion of the accumulated flow consumption in the total flow consumption at the current accumulated minutes is calculated according to the accumulated minutes t by using the operation formula “-8.009*10 -10 t 3 +2.01*10 -6 t 2 -8.081*10 -4 t+0.1877”, and the calculated result is taken as the flow consumption proportion corresponding to the current accumulated minutes output by the first curve fitting result.

[0322] In step 1330, the second curve fitting result is determined based on the second proportions of each sampling time point of the non-target time day by using the least square method.

[0323] In the implementation of this embodiment, the specific process of step 1330 is similar to the specific process of step 1320 described above. For the sake of brevity, no longer be described.

[0324] Wherein, the second curve fitting result can be represented as formula (4) shown:

[0325]

[0326] Wherein, F weekend (t) is the predicted proportion of the flow consumption at each accumulated minutes of the weekend in the total flow consumption.

[0327] The embodiment has the advantages that, considering the differences in flow consumption in different natural days and different times of the natural days, the flow consumption distribution fitting is performed on the flow consumption data of multiple natural days, the flow consumption distribution results of working days and non-working days are fitted respectively, and in the flow consumption distribution results of working days and non-working days, the flow consumption distribution results are divided into segmented functions according to different time periods, so that the fitting effect of the flow consumption distribution is improved. Further, the current flow consumption situation is determined by using the target curve fitting result according to the current time day and the current cumulative minutes, and the current expected consumption resource of the target advertisement is allocated according to the flow consumption situation, so that the allocation of the resource constraint data is as close as possible to the real flow consumption situation, and the conversion amount of the target advertisement is as large as possible.

[0328] In one embodiment, the target curve fitting result includes multiple candidate time intervals and a curve equation corresponding to each candidate time interval.

[0329] The candidate time interval is used to divide all minute points of a natural day into multiple intervals, and each interval is defined as an interval corresponding to different cumulative minutes. For example, a natural day is divided into three candidate time intervals, candidate time interval 1 corresponds to cumulative minutes (0, 360]; candidate time interval 2 corresponds to cumulative minutes (360, 840]; and candidate time interval 3 corresponds to cumulative minutes (840, 1440].

[0330] The curve equation is used to indicate the mathematical operation form of the flow consumption situation corresponding to different cumulative minutes.

[0331] Please refer to Figure 14 In the embodiment, step 1230 specifically includes but is not limited to steps 1410-1420.

[0332] Step 1410, determining a target time interval in multiple candidate time intervals based on the cumulative time data;

[0333] Step 1420, substituting the cumulative time data into the curve equation corresponding to the target time interval to obtain the predicted flow consumption data.

[0334] The steps 1410-1420 are described in detail as follows.

[0335] In step 1410, a target time interval is determined in multiple candidate time intervals based on the cumulative time data.

[0336] The target time interval is used to indicate the candidate time interval in which the cumulative time data is located.

[0337] In the embodiment, first, the accumulated time data is compared with the interval end point values of each candidate time interval. Then, the candidate time interval with one interval end point value not greater than the accumulated time data and the other interval end point value not less than the accumulated time data is determined as the candidate time interval where the accumulated time data is located. Finally, the candidate time interval where the accumulated time data is located is determined as the target time interval.

[0338] In step 1420, the accumulated time data is substituted into the curve equation corresponding to the target time interval to obtain the predicted traffic consumption data.

[0339] In the embodiment, first, the accumulated time data is substituted into the curve equation corresponding to the target time interval to obtain the output result of the curve equation. Then, if there is no numerical comparison link in the target curve fitting result, the output result of the curve equation is directly taken as the predicted traffic consumption data at the current time. If there is a numerical comparison link in the target curve fitting result, the output result of the curve equation and the predetermined numerical value are compared according to the comparison rule of the numerical comparison link, and the output result of the curve equation and the predetermined numerical value that meet the comparison value requirement are taken as the predicted traffic consumption data at the current time.

[0340] The embodiment has the advantage that the accumulated traffic consumption at the current minute of the current time day accounts for the total traffic consumption of the whole day is calculated by the fitting result of the traffic consumption distribution through mathematical operation, which can improve the prediction accuracy and prediction efficiency of the traffic consumption at each minute point at the minute level.

[0341] Detailed description of step 350

[0342] In step 350, the initial bid is updated based on the comparison between the actual consumed resource data and the expected consumed resource data, and the step of determining the expected consumed resource data of the target advertisement based on the preset traffic consumption data and the predetermined resource constraint data is returned until the target advertisement reaches the predetermined exploration condition, and the target bid of the target advertisement is determined based on the updated initial bid.

[0343] Please refer to Figure 15 In one embodiment, the specific process of updating the initial bid based on the comparison between the actual consumed resource data and the expected consumed resource data includes but is not limited to the following steps 1510-1520:

[0344] Step 1510, determining the bid adjustment parameter of the target advertisement based on the comparison between the actual consumed resource data and the expected consumed resource data;

[0345] Step 1520, updating the initial bid based on the bid adjustment parameter.

[0346] The steps 1510-1520 are described in detail as follows.

[0347] In step 1510, a bid adjustment parameter of the target advertisement is determined based on a comparison between the actual consumed resource data and the expected consumed resource data.

[0348] The bid adjustment parameter is used to indicate an adjustment magnitude of the initial bid of the target advertisement according to the comparison between the actual consumed resource data and the expected consumed resource data.

[0349] In the embodiment, first, the actual consumed resource data and the expected consumed resource data of the target advertisement at the current minute number (the minute number of the current natural day) are compared. If the actual consumed resource data is greater than the expected consumed resource data, it indicates that the resource consumption of the target advertisement at the current minute number is too fast, and the resource consumption speed of the target advertisement needs to be slowed down. Based on this, the bid adjustment parameter is set to a parameter that can reduce the initial bid. If the actual consumed resource data is less than the expected consumed resource data, it indicates that the resource consumption of the target advertisement at the current minute number is too slow, and the resource consumption speed of the target advertisement needs to be accelerated. Based on this, the bid adjustment parameter is set to a parameter that can increase the initial bid.

[0350] In step 1520, the initial bid is updated based on the bid adjustment parameter.

[0351] In the embodiment, when the initial bid is updated based on the bid adjustment parameter, first, the bid adjustment parameter and 1 are summed to obtain a third sum, wherein the third sum is used to indicate the size relationship between the updated initial bid at the current minute point and the initial bid at the previous minute point of the current minute point. Then, the third sum and the initial bid are multiplied, and the product of the third sum and the initial bid is taken as the updated initial bid at the current minute point.

[0352] Wherein, the updated initial bid at each minute point can be represented as formula (5):

[0353] bid t = (1 + λ t ) * bid t-1 of formula (5);

[0354] Wherein, bid t refers to the updated initial bid at the current minute point, bid t-1 refers to the updated initial bid at the previous minute point of the current minute point. λ t refers to the bid adjustment parameter at the current minute point t.

[0355] The embodiment has the advantages that, according to comparison of the actual consumed resource data and the expected consumed resource data, when the actual resource consumption of the target advertisement is too fast, the bid adjustment parameter is set to a value that can slow down the resource consumption speed of the target advertisement; when the actual resource consumption of the target advertisement is too slow, the bid adjustment parameter is set to a value that can accelerate the resource consumption speed of the target advertisement, the bid adjustment parameter of the target advertisement can be flexibly controlled, and the initial bid is dynamically updated based on the bid adjustment parameter, so that the adjusted advertisement bid can make the actual resource consumption and the expected resource consumption of the target advertisement as consistent as possible, and the effect of maximizing the conversion quantity of the target advertisement under the condition of consuming the expected resource quantity as much as possible is achieved.

[0356] Please refer to Figure 16 In one embodiment, step 1510 specifically includes but is not limited to steps 1610-1630:

[0357] Step 1610, determining a first value based on a quotient of the expected consumed resource data and the actual consumed resource data;

[0358] Step 1620, determining a smaller value as an intermediate parameter from the first value, and a first predetermined value;

[0359] Step 1630, determining a larger value as the bid adjustment parameter from the intermediate parameter, and a second predetermined value.

[0360] The steps 1610-1630 are described in detail as follows.

[0361] In step 1610, a first value is determined based on a quotient of the expected consumed resource data and the actual consumed resource data.

[0362] In the implementation of the embodiment, first, the expected consumed resource data is divided by the actual consumed resource data to obtain the quotient of the expected consumed resource data and the actual consumed resource data; then, the quotient of the expected consumed resource data and the actual consumed resource data is determined as the first value. When the first value is greater than 1, it indicates that the expected consumed resource data is greater than the actual consumed resource data, and the target advertisement consumes resources too slowly. When the first value is less than 1, it indicates that the expected consumed resource data is less than the actual consumed resource data, and the target advertisement consumes resources too fast.

[0363] In step 1620, a smaller value is determined as an intermediate parameter from the first value, and a first predetermined value.

[0364] In the implementation of the embodiment, the first value and the first predetermined value are compared, and a smaller value is obtained between the first value and the first predetermined value to obtain the intermediate parameter.

[0365] In step 1630, among the intermediate parameter, and the second predetermined value, the larger value is determined as the bid adjustment parameter.

[0366] The first predetermined value is larger than the second predetermined value, and the first predetermined value and the second predetermined value are used to constrain the change range of the bid adjustment parameter.

[0367] In the implementation of this embodiment, the intermediate parameter, and the second predetermined value are compared, and the larger one between the intermediate parameter, and the second predetermined value is taken as the bid adjustment parameter.

[0368] Wherein, the bid adjustment parameter of the target advertisement at the current minute point t can be represented as formula (6) shows:

[0369]

[0370] Wherein, λ t The value range of is [-0.01, 0.01]; expectCost t is the expected consumption resource data of the target advertisement at the current minute t; realCost t is the actual consumption resource data of the target advertisement at the current minute t.

[0371] The above formula (6) represents: for the bid adjustment parameter of the target advertisement at the current minute point t, first take the smaller one between the quotient of the expected consumption resource data and the actual consumption resource data, and 0.01, to get the intermediate parameter Then take the larger one between the intermediate parameter and -0.01 as the bid adjustment parameter λ t .

[0372] The advantage of this embodiment is that the difference between the actual consumption resource data and the expected consumption resource data is judged according to the quotient of the expected consumption resource data and the expected actual consumption resource data, and the bid adjustment parameter is dynamically adjusted at each minute point according to the difference between the two, so as to dynamically adjust the bid size of the target advertisement at each minute point. Further, the value range of the bid adjustment parameter is also constrained, so that the bid adjustment parameter fluctuates between ±0.0.1, which can effectively avoid the fluctuation of the calculated bid adjustment parameter being too large, and improve the accuracy and stability of the bid adjustment parameter.

[0373] Please refer to Figure 17 In one embodiment, the specific process of determining the target bid of the target advertisement based on the updated initial bid includes but is not limited to the following steps 1710-1730:

[0374] Step 1710, determine the upper limit parameter of the bid of the target advertisement;

[0375] Step 1720, determining the maximum bid of the target advertisement based on the product of the bid upper limit parameter and the predicted associated resource data;

[0376] Step 1730, updating the initial bid based on the bid adjustment parameter.

[0377] The steps 1710-1730 are described in detail as follows.

[0378] In step 1710, the bid upper limit parameter of the target advertisement is determined.

[0379] The bid upper limit parameter is used to indicate the maximum interaction resource amount (maximum bid) that the advertiser can provide in the advertisement bidding opportunity.

[0380] In the embodiment, the bid upper limit parameter of the target advertisement is generally preset. Based on this, the bid upper limit parameter set for the target advertisement can be directly obtained under the authorized permission.

[0381] The bid upper limit parameter of the target advertisement can be determined according to the delivery scenario, the advertisement type, etc. of the target advertisement, without limitation.

[0382] In step 1720, the maximum bid of the target advertisement is determined based on the product of the bid upper limit parameter and the predicted associated resource data.

[0383] In the embodiment, the bid upper limit parameter and the predicted associated resource data are multiplied, and the predicted associated resource data is amplified to appropriately increase the estimated interaction resource upper limit of the target advertisement, so as to obtain the maximum bid of the target advertisement.

[0384] In step 1730, the smaller one of the updated initial bid and the maximum bid is determined as the target bid of the target advertisement.

[0385] In the embodiment, first, the updated initial bid and the maximum bid of the target advertisement at the current minute point t are compared. Then, the smaller one of the updated initial bid and the maximum bid is taken as the maximum bid of the target advertisement, so that the target bid is not greater than the estimated interaction resource upper limit of the target advertisement when the target bid is dynamically adjusted.

[0386] The target bid of the target advertisement at the current minute point t can be represented as formula (7):

[0387] bid t =min(bid t ,γ*CPA) formula (7);

[0388] min(bid tbid t is the updated initial bid; γ is a bid cap parameter; CPA is the predicted associated resource data; γ*CPA refers to the maximum bid of the target advertisement. min(bid t , γ*CPA) means taking the smaller one between the updated initial bid and the maximum bid. The value of γ ranges from (1, 2]. In order to improve the accuracy of the target bid, γ can be taken as 1.4.

[0389] The benefit of this embodiment is that the updated initial bid is truncated by using the bid cap parameter, so that the target bid of the target advertisement is not greater than the maximum bid, which can effectively reduce the risk that the target bid of the target advertisement is too high and the target advertisement consumes resources too quickly, and can greatly improve the conversion effect of the target advertisement.

[0390] Implementation details of the advertisement processing method in the embodiments of the present disclosure

[0391] The following refers to Figure 18 , and the implementation details of the advertisement processing method of the embodiment of the present disclosure are exemplarily and specifically illustrated.

[0392] As Figure 18 indicated, when bidding for the target advertisement at the advertisement opportunity, first, the initial bid of the target advertisement is determined by using the predicted associated resource module of the advertisement system based on the advertisement processing method of the present disclosure, wherein the predicted associated resource module will predict the predicted associated resource data of the target advertisement in the order of “the product SKU of the product to which the advertisement belongs→the product SPU of the product to which the advertisement belongs→the third-level category of the product to which the advertisement belongs→the second-level category of the product to which the advertisement belongs”, and the specific implementation process is similar to steps 310-320 described above.

[0393] Further, the resource constraint data distribution module of the advertisement system is used to fit the curve of the traffic consumption amount changing with time, to obtain the first curve fitting result on weekdays and the second curve fitting result on weekends, and the specific implementation process is similar to steps 1310-1330 described above.

[0394] Then, the bid exploration module of the advertisement system is used to determine, for each minute point of the current time day, whether the target advertisement has been exposed at the initial bid of the target advertisement at the current minute point. If the target advertisement is not exposed, the bid of the target advertisement is raised, and the specific implementation process is similar to step 930 described above. If the target advertisement is exposed, it is determined that the current initial bid of the target advertisement meets the requirements, and the specific implementation process is similar to step 920 described above.

[0395] Further, when it is determined that the target advertisement has exposure, it is judged whether the resource consumption of the target advertisement is too slow, if the resource consumption of the target advertisement is too slow, the bid of the target advertisement is raised, if the resource consumption of the target advertisement is too fast, the bid of the target advertisement is lowered, if the resource consumption of the target advertisement does not exist too fast or too slow, the bid of the target advertisement is not adjusted, and the specific implementation process is similar to steps 1510-1520 described above. After the bid of the target advertisement is lowered, the adjusted bid is directly output as the target bid of the target advertisement at the current minute point.

[0396] When the bid of the target advertisement is raised, it is also necessary to judge whether the bid of the target advertisement exceeds the maximum bid (the upper limit of the estimated cost). If it is determined that the bid of the target advertisement does not exceed the maximum bid, the bid of the target advertisement is not adjusted, and the current bid is output as the target bid of the target advertisement at the current minute point; if it is determined that the bid of the target advertisement exceeds the maximum bid, the maximum bid is taken as the target bid of the target advertisement at the current minute point, and the specific implementation process is similar to steps 1710-1730 described above. In order to save space, it will not be repeated here.

[0397] Device and equipment description in the embodiments of the present disclosure

[0398] It can be understood that although each step in each of the above flowcharts is displayed in sequence according to the arrow representation, these steps are not necessarily executed in the order represented by the arrow. Unless otherwise specified in this embodiment, the execution of these steps has no strict order limitation, and these steps can be executed in other order. Moreover, at least part of the steps in the above flowcharts can include multiple steps or multiple stages, which are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily sequential, but can be alternately executed with at least part of other steps or steps or stages in other steps.

[0399] It should be noted that in each specific embodiment of the present application, when it is necessary to process related data such as target object attribute information or attribute information set related to the characteristics of the target object, the permission or consent of the target object will be obtained first, and the collection, use and processing of these data will comply with relevant laws, regulations and standards. In addition, when the target object attribute information is required by the embodiments of the present application, the separate permission or separate consent of the target object will be obtained through a pop-up window or by jumping to a confirmation page, and after obtaining the separate permission or separate consent of the target object, the necessary target object related data required for the normal operation of the embodiments of the present application will be obtained.

[0400] Figure 19A structural schematic diagram of an advertisement processing apparatus 1900 is provided for embodiments of the present disclosure. The advertisement processing apparatus 1900 comprises:

[0401] An acquisition unit 1910 is configured to acquire target associated resource data of a target advertisement, wherein the target associated resource data is used to indicate an input resource amount associated with the target advertisement in a target dimension;

[0402] A first determination unit 1920 is configured to determine predicted associated resource data of the target advertisement based on the target associated resource data and predetermined resource constraint data of the target advertisement, and determine an initial bid of the target advertisement based on the predicted associated resource data;

[0403] A second determination unit 1930 is configured to perform bid exploration based on the initial bid, and determine actual consumed resource data of the target advertisement;

[0404] A third determination unit 1940 is configured to, if it is determined that the actual consumed resource data meets a predetermined condition, determine expected consumed resource data of the target advertisement based on pre-acquired traffic consumption data and the predetermined resource constraint data;

[0405] An updating unit 1950 is configured to update the initial bid based on a comparison between the actual consumed resource data and the expected consumed resource data, and return to the step of determining the expected consumed resource data of the target advertisement based on the pre-set traffic consumption data and the predetermined resource constraint data until the target advertisement meets a predetermined exploration condition, and determine a target bid of the target advertisement based on the updated initial bid.

[0406] Optionally, the acquisition unit 1910 is configured to:

[0407] acquire target exposure data of the target advertisement in the target dimension;

[0408] if it is determined that the target exposure data does not reach a predetermined exposure threshold, acquire reference exposure data of the target advertisement in a higher-level dimension of the target dimension;

[0409] replace the target dimension with the higher-level dimension of the target dimension, and replace the target exposure data with the reference exposure data, and return to the step of acquiring the reference exposure data of the target advertisement in the higher-level dimension of the target dimension if it is determined that the target exposure data does not reach the predetermined exposure threshold, until the target exposure data reaches the predetermined exposure threshold, and acquire an input resource mean value associated with the target dimension of the target advertisement;

[0410] determine the target associated resource data based on the input resource mean value.

[0411] Optionally, the first determination unit 1920 is configured to:

[0412] determine a number of associated items generated by the target advertisement in a target dimension in a target time day;

[0413] determine predicted associated resource data based on the number of associated items, target associated resource data, and predetermined resource constraint data.

[0414] Optionally, the determining of the predicted associated resource data based on the number of associated items, target associated resource data, and predetermined resource constraint data comprises:

[0415] determine a first product based on a product of the number of associated items and the target associated resource data;

[0416] perform logarithmic scaling on a comparison result of the predetermined resource constraint data and the first product to obtain a logarithmic scaling result;

[0417] perform scaling processing on the logarithmic scaling result and a first sum of 1 and the target associated resource data to obtain the predicted associated resource data.

[0418] Optionally, the first determining unit 1920 is configured to:

[0419] determine an exploration factor of the target advertisement based on a starting speed of the target advertisement and associated input resources;

[0420] determine an initial bid based on a product of an inverse of the exploration factor and the predicted associated resource data.

[0421] Optionally, the third determining unit 1940 is configured to:

[0422] perform fitting processing on the traffic consumption data to obtain a first curve fitting result indicating a traffic consumption distribution of the target time day and a second curve fitting result indicating a traffic consumption distribution of a non-target time day;

[0423] determine a current time day and cumulative time data of the current time day;

[0424] determine a target curve fitting result from the first curve fitting result and the second curve fitting result based on the current time day;

[0425] determine predicted traffic consumption data based on the cumulative time data and the target curve fitting result;

[0426] determine expected consumption resource data of the target advertisement based on a product of the predicted traffic consumption data and predetermined resource constraint data.

[0427] Optionally, the performing of the fitting processing on the traffic consumption data to obtain the first curve fitting result indicating the traffic consumption distribution of the target time day and the second curve fitting result indicating the traffic consumption distribution of the non-target time day comprises:

[0428] determine a first proportion of the cumulative flow consumption amount of each sampling time point of the target time day in a total flow consumption amount of the target time day, and determine a second proportion of the cumulative flow consumption amount of each sampling time point of the non-target time day in a total flow consumption amount of the non-target time day based on the flow consumption data;

[0429] determine a first curve fitting result by using a least square method based on the first proportion of each sampling time point of the target time day;

[0430] determine a second curve fitting result by using a least square method based on the second proportion of each sampling time point of the non-target time day.

[0431] Optionally, the target curve fitting result includes a plurality of candidate time intervals and a curve equation corresponding to each candidate time interval;

[0432] determine predicted flow consumption data based on the cumulative time data and the target curve fitting result, including:

[0433] determine a target time interval from the plurality of candidate time intervals based on the cumulative time data;

[0434] substitute the cumulative time data into the curve equation corresponding to the target time interval to obtain the predicted flow consumption data.

[0435] Optionally, the updating unit 1950 includes:

[0436] a comparison module (not shown) configured to determine a bid adjustment parameter of the target advertisement based on a comparison between the actual consumed resource data and the expected consumed resource data;

[0437] an updating module (not shown) configured to update the initial bid based on the bid adjustment parameter.

[0438] Optionally, the comparison module (not shown) is configured to:

[0439] determine a first value based on a quotient of the expected consumed resource data and the actual consumed resource data;

[0440] determine a smaller value as an intermediate parameter from among the first value, and a first predetermined value;

[0441] determine a larger value as the bid adjustment parameter from among the intermediate parameter and a second predetermined value, wherein the first predetermined value is greater than the second predetermined value.

[0442] Optionally, the updating unit 1950 is configured to:

[0443] determine an upper limit parameter of the bid of the target advertisement;

[0444] determine the maximum bid of the target advertisement based on a product of the bid upper limit parameter and the predicted associated resource data;

[0445] determine the target bid of the target advertisement as the smaller one of the updated initial bid and the maximum bid.

[0446] Optionally, the advertisement processing apparatus 1900 further comprises a judging unit (not shown), configured to:

[0447] determine whether the actual consumed resource data is non-zero data;

[0448] if it is determined that the actual consumed resource data is non-zero data, determine that the actual consumed resource data satisfies a predetermined condition;

[0449] if it is determined that the actual consumed resource data is not non-zero data, perform bid adjustment on the initial bid based on a predetermined bid adjustment factor until the actual consumed resource data is non-zero data, and replace the initial bid with the adjusted initial bid.

[0450] Optionally, performing bid adjustment on the initial bid based on the predetermined bid adjustment factor comprises:

[0451] determining a second sum of the predetermined bid adjustment factor and 1;

[0452] determining the adjusted initial bid based on a product of the second sum and the initial bid.

[0453] Optionally, the predetermined bid adjustment factor is determined by:

[0454] determining a winning difficulty level of the target advertisement and a delivery scenario;

[0455] determining a first score based on the winning difficulty level;

[0456] determining a second score based on the delivery scenario;

[0457] determining the predetermined bid adjustment factor based on the first score and the second score.

[0458] Reference Figure 20 , Figure 20 The structure block diagram of a part of a terminal for implementing the advertisement processing method of the embodiments of the present disclosure includes a radio frequency (RF) circuit 2010, a memory 2015, an input unit 2030, a display unit 2040, a sensor 2050, an audio circuit 2060, a wireless fidelity (WiFi) module 2070, a processor 2080, and a power supply 2090, and the like. Those skilled in the art can understand that the terminal can further include other components, and the components are not limited to the above-mentioned components. Figure 20The terminal structure shown does not constitute a limitation on the mobile phone or computer, and can include more or fewer components than shown, or combine certain components, or different component arrangements.

[0459] The RF circuit 2010 can be used for receiving and sending signals in the process of information or call, in particular, receiving the downlink information of the base station and processing by the processor 2080; in addition, sending the uplink data to the base station.

[0460] The memory 2015 can be used for storing software programs and modules, and the processor 2080 executes various function applications and data processing of the object terminal by running the software programs and modules stored in the memory 2015.

[0461] The input unit 2030 can be used for receiving input digital or character information, and generating key signal input related to the setting and function control of the object terminal. Specifically, the input unit 2030 can include a touch panel 2031 and other input devices 2032.

[0462] The display unit 2040 can be used for displaying input information or provided information and various menus of the object terminal. The display unit 2040 can include a display panel 2041.

[0463] The audio circuit 2060, the speaker 2061, and the microphone 2062 can provide an audio interface.

[0464] In the embodiment, the processor 2080 included in the terminal can execute the advertisement processing method of the previous embodiment.

[0465] The terminal of the embodiments of the present disclosure includes but is not limited to a mobile phone, a computer, a smart voice interaction device, a smart home appliance, a vehicle-mounted terminal, an aircraft, etc. The embodiments of the present application can be applied to various scenarios, including but not limited to advertisement delivery, big data, data storage, information technology, etc.

[0466] Figure 21A structure block diagram of a server for implementing the advertisement processing method of the embodiments of the present disclosure. The server can be quite different due to different configurations or performances, and can include one or more central processing units (CPUs) 2122 (e.g., one or more processors) and a memory 2132, one or more storage media 2130 (e.g., one or more mass storage devices) storing applications 2142 or data 2144. Among them, the memory 2132 and the storage medium 2130 can be temporary storage or persistent storage. The programs stored in the storage medium 2130 can include one or more modules (not shown in the figure), each of which can include a series of instruction operations in the server. Further, the central processing unit 2122 can be configured to communicate with the storage medium 2130 and execute the series of instruction operations in the storage medium 2130 on the server.

[0467] The server can also include one or more power supplies 2126, one or more wired or wireless network interfaces 2150, one or more input / output interfaces 2158, and / or one or more operating systems 2141, such as Windows ServerTM, Mac OS XTM, UnixTM, LinuxTM, FreeBSDTM, etc.

[0468] The central processing unit 2122 in the server can be used to execute the advertisement processing method of the embodiments of the present disclosure.

[0469] The embodiments of the present disclosure also provide a computer-readable storage medium for storing program code for executing the advertisement processing method of the various embodiments described above.

[0470] The embodiments of the present disclosure also provide a computer program product including a computer program. The processor of the computer device reads and executes the computer program, so that the computer device executes the advertisement processing method as described above.

[0471] The terms "first", "second", "third", "fourth", and the like in the description of the disclosure and the above drawings, if any, are used to distinguish similar objects, and do not necessarily have to be used to describe a particular order or sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments of the disclosure described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "include" and "contain" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device containing a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but can include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0472] It should be understood that in the present disclosure, "at least one" refers to one or more, and "multiple" refers to two or more. "And / or" is used to describe the relationship between the associated objects, which means that there can be three relationships, for example, "A and / or B" can represent three cases: only A exists, only B exists, and A and B exist at the same time, where A and B can be singular or plural. The character " / " generally represents an "or" relationship between the associated objects. "At least one of the following" or similar expressions means any combination of these items, including any combination of single or multiple items. For example, at least one of a, b or c can represent: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be single or multiple.

[0473] It should be understood that in the description of the embodiments of the present disclosure, the meaning of multiple (or multiple items) is two or more, greater than, less than, more than, etc. are not included in the number, above, below, etc. are understood to include the number.

[0474] In several embodiments provided by the present disclosure, it should be understood that the disclosed system, device and method can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of units is only a logical function division, and actual implementation can have another division manner, for example, multiple units or components can be combined or integrated into another system, or some features can be omitted or not executed. In addition, the coupling or direct coupling or communication connection between the displayed or discussed objects can be indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0475] The units described as separate components may or may not be physically separate, and the components displayed as units may or may not be physical units, that is, may be located in one place, or may be distributed to multiple network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the embodiment scheme.

[0476] In addition, each functional unit in various embodiments of the present disclosure can be integrated into one processing unit, or each unit can exist physically, or two or more units can be integrated into one unit. The integrated unit can be realized in the form of hardware or in the form of a software functional unit.

[0477] If the integrated unit is realized in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the present disclosure essentially or the part that contributes to the prior art or the whole or part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium, and includes a number of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the various embodiments of the method of the present disclosure. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and various media that can store program codes.

[0478] It should also be appreciated that the various embodiments provided by the present disclosure can be combined in any way to achieve different technical effects.

[0479] The above is a specific description of the embodiments of the present disclosure, but the present disclosure is not limited to the above embodiments. Those skilled in the art can make various equivalent modifications or replacements without departing from the spirit of the present disclosure, and these equivalent modifications or replacements are included in the scope defined by the claims of the present disclosure.

Claims

1. An advertisement processing method characterized by comprising: The method comprises: acquiring target associated resource data of a target advertisement, wherein the target associated resource data is used to indicate an input resource amount associated with the target advertisement in a target dimension; determining predicted associated resource data of the target advertisement based on the target associated resource data and predetermined resource constraint data of the target advertisement, and determining an initial bid of the target advertisement based on the predicted associated resource data; performing bid exploration based on the initial bid, and determining actual consumed resource data of the target advertisement; if it is determined that the actual consumed resource data meets a predetermined condition, determining expected consumed resource data of the target advertisement based on pre-acquired traffic consumption data and the predetermined resource constraint data; updating the initial bid based on a comparison between the actual consumed resource data and the expected consumed resource data, and returning to the step of determining the expected consumed resource data of the target advertisement based on the pre-set traffic consumption data and the predetermined resource constraint data until the target advertisement reaches a predetermined exploration condition, and determining a target bid of the target advertisement based on the updated initial bid.

2. The advertisement processing method according to claim 1, characterized by, The acquiring of the target associated resource data of the target advertisement comprises: acquiring target exposure data of the target advertisement in a target dimension; if it is determined that the target exposure data does not reach a predetermined exposure threshold, acquiring reference exposure data of the target advertisement in a higher-level dimension of the target dimension; replacing the target dimension with the higher-level dimension of the target dimension and replacing the target exposure data with the reference exposure data, and returning to the step of acquiring the reference exposure data of the target advertisement in the higher-level dimension of the target dimension if it is determined that the target exposure data does not reach the predetermined exposure threshold until the target exposure data reaches the predetermined exposure threshold, and acquiring an input resource mean value associated with the target dimension of the target advertisement; determining the target associated resource data based on the input resource mean value.

3. The advertisement processing method of claim 1, wherein, The determining of the predicted associated resource data of the target advertisement based on the target associated resource data and the predetermined resource constraint data of the target advertisement comprises: determining a number of associated items generated by the target advertisement in a target dimension in a time day; determining the predicted associated resource data based on the number of associated items, the target associated resource data, and the predetermined resource constraint data.

4. The advertisement processing method according to claim 3, characterized by, The determining of the predicted associated resource data based on the number of associated items, the target associated resource data, and the predetermined resource constraint data comprises: determining a first product based on a product of the number of associated items and the target associated resource data; performing logarithmic scaling on a comparison result of the predetermined resource constraint data and the first product to obtain a logarithmic scaling result; performing scaling processing on the logarithmic scaling result and 1 and the target associated resource data to obtain the predicted associated resource data.

5. The advertisement processing method of claim 1, wherein, The determining of the initial bid of the target advertisement based on the predicted associated resource data comprises: determining an exploration factor of the target advertisement based on a starting speed of the target advertisement and associated input resources; determine the initial bid based on a product of a reciprocal of the exploration factor and the predicted associated resource data.

6. The advertisement processing method of claim 1, wherein, The determining the expected consumption resource data of the target advertisement based on the pre-acquired traffic consumption data and the predetermined resource constraint data comprises: performing fitting processing on the traffic consumption data to obtain a first curve fitting result for indicating a traffic consumption distribution of a target time day and a second curve fitting result for indicating a traffic consumption distribution of a non-target time day; determining a current time day and accumulated time data of the current time day; determining a target curve fitting result from the first curve fitting result and the second curve fitting result based on the current time day; determining predicted traffic consumption data based on the accumulated time data and the target curve fitting result; determining the expected consumption resource data of the target advertisement based on a product of the predicted traffic consumption data and the predetermined resource constraint data.

7. The advertisement processing method according to claim 6, wherein The performing fitting processing on the traffic consumption data to obtain a first curve fitting result for indicating a traffic consumption distribution of a target time day and a second curve fitting result for indicating a traffic consumption distribution of a non-target time day comprises: determining a first proportion of a cumulative traffic consumption amount of each sampling time point of the target time day in a total amount of traffic consumption of the target time day and a second proportion of a cumulative traffic consumption amount of each sampling time point of the non-target time day in a total amount of traffic consumption of the non-target time day based on the traffic consumption data; determining the first curve fitting result based on the first proportion of each sampling time point of the target time day by using a least square method; determining the second curve fitting result based on the second proportion of each sampling time point of the non-target time day by using the least square method.

8. The advertisement processing method of claim 1, wherein, The updating the initial bid based on the comparison between the actual consumption resource data and the expected consumption resource data comprises: determining a bid adjustment parameter of the target advertisement based on the comparison between the actual consumption resource data and the expected consumption resource data; updating the initial bid based on the bid adjustment parameter.

9. The advertisement processing method according to claim 8, wherein The determining a bid adjustment parameter of the target advertisement based on the comparison between the actual consumption resource data and the expected consumption resource data comprises: determining a first value based on a quotient of the expected consumption resource data and the actual consumption resource data; determining a smaller one as an intermediate parameter from the first value and a first predetermined value; determining a larger one as the bid adjustment parameter from the intermediate parameter and a second predetermined value, wherein the first predetermined value is larger than the second predetermined value.

10. The advertisement processing method of claim 1, wherein, The determining a target bid of the target advertisement based on the updated initial bid comprises: determining an upper limit parameter of a bid of the target advertisement; determining a maximum bid of the target advertisement based on a product of the upper limit parameter of the bid and the predicted associated resource data; determining a smaller one as the target bid of the target advertisement from the updated initial bid and the maximum bid.

11. The advertisement processing method of claim 1, wherein, After the actual consumed resource data of the target advertisement is determined, the method further comprises: determining whether the actual consumed resource data is non-zero data; if it is determined that the actual consumed resource data is non-zero data, determining that the actual consumed resource data meets a predetermined condition; if it is determined that the actual consumed resource data is not non-zero data, adjusting the initial bid based on a predetermined bid adjustment factor until the actual consumed resource data is non-zero data, and replacing the initial bid with an adjusted initial bid.

12. The advertisement processing method according to claim 11, wherein The adjusting of the initial bid based on a predetermined bid adjustment factor comprises: determining a second sum of the predetermined bid adjustment factor and 1; determining the adjusted initial bid based on a product of the second sum and the initial bid.

13. An advertisement processing apparatus characterized by comprising: The apparatus comprises: an obtaining unit configured to obtain target associated resource data of a target advertisement, wherein the target associated resource data is used to indicate an amount of input resources associated with the target advertisement in a target dimension; a first determining unit configured to determine predicted associated resource data of the target advertisement based on the target associated resource data and predetermined resource constraint data of the target advertisement, and determine an initial bid of the target advertisement based on the predicted associated resource data; a second determining unit configured to perform bid exploration based on the initial bid, and determine actual consumed resource data of the target advertisement; a third determining unit configured to, if it is determined that the actual consumed resource data meets a predetermined condition, determine expected consumed resource data of the target advertisement based on previously obtained traffic consumption data and the predetermined resource constraint data; an updating unit configured to update the initial bid based on a comparison between the actual consumed resource data and the expected consumed resource data, and return to the step of determining the expected consumed resource data of the target advertisement based on the predetermined traffic consumption data and the predetermined resource constraint data until the target advertisement meets a predetermined exploration condition, and determine a target bid of the target advertisement based on the updated initial bid.

14. An electronic device, comprising a memory and a processor, the memory storing a computer program, characterized in that, The processor executes the computer program to implement the advertisement processing method of any one of claims 1 to 12.

15. A computer readable storage medium, the storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the advertisement processing method of any one of claims 1 to 12.

16. A computer program product comprising a computer program, which is read and executed by a processor of a computer device, so that the computer device performs the advertisement processing method of any one of claims 1 to 12.

Images