Budget allocation method, computing device, storage medium and program product

Abstract

The embodiment of the invention provides a budget allocation method, computing equipment, a storage medium and a program product. Wherein the method aims at a plurality of advertisement plans of an advertiser; estimating at least one estimated income value generated under the condition that any one advertisement plan is allocated to any one candidate budget; determining candidate budgets respectively matched with the plurality of advertisement plans by taking at least one estimated income value respectively corresponding to the plurality of advertisement plans to meet the income value requirement as an optimization target; and taking the candidate budgets respectively matched with the plurality of advertisement plans as target budgets respectively corresponding to the plurality of advertisement plans. According to the technical scheme provided by the embodiment of the invention, the advertisement putting effect is improved.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of computer technology, and in particular to a budget allocation method, a computing device, a storage medium, and a program product. Background Art

[0002] External advertising refers to the promotional behavior of placing advertising plans for advertising objects to external advertising delivery systems such as various media platforms, and paying related fees. Advertising plans are mainly composed of one or more advertising contents generated by advertising objects, target users, and advertising materials. After allocating a budget for the advertising plan, the advertising delivery system can perform promotion operations based on the advertising plan under budget constraints.

[0003] Currently, advertisers usually set a fixed budget for each advertising plan during the delivery cycle, which results in that advertising plans with better performance cannot continue to be displayed during the delivery cycle due to budget constraints, and poorly performing ads or campaigns may waste budgets, ultimately leading to low delivery efficiency and affecting the effectiveness of advertising delivery. Therefore, how to reasonably allocate budgets has become a technical problem that needs to be solved. Summary of the invention

[0004] The embodiments of the present application provide a budget allocation method, a computing device, a storage medium, and a program product to solve the problem of poor advertising delivery effect in the prior art.

[0005] In a first aspect, an embodiment of the present application provides a budget allocation method, comprising:

[0006] Identify multiple advertising plans;

[0007] estimating at least one estimated revenue value generated by allocating any candidate budget to any advertising plan;

[0008] Taking at least one estimated revenue value corresponding to each of the multiple advertising plans satisfying a revenue value requirement as an optimization goal, determining candidate budgets that match each of the multiple advertising plans;

[0009] The candidate budgets respectively matched to the multiple advertising plans are used as the target budgets respectively corresponding to the multiple advertising plans.

[0010] Optionally, the estimating at least one estimated revenue value generated when any advertising plan is allocated to any candidate budget includes:

[0011] Using at least one revenue estimation model, estimating at least one estimated revenue value generated when any advertising plan is allocated with any candidate budget;

[0012] The revenue estimation model is obtained by training based on a sample advertising plan, a sample budget corresponding to the sample advertising plan, and a sample revenue value.

[0013] Optionally, the estimating at least one estimated revenue value generated when any advertising plan is allocated any candidate budget by using at least one revenue estimation model includes:

[0014] Determining at least one advertisement characteristic corresponding to any advertisement plan;

[0015] The at least one advertisement feature and any candidate budget are input into at least one revenue estimation model to obtain at least one estimated revenue value generated when the candidate budget is allocated to the advertisement plan.

[0016] Optionally, determining at least one advertisement feature corresponding to any advertisement plan includes:

[0017] Determine at least one average revenue value generated by any advertising program over a historical time period;

[0018] One or more of the at least one average revenue value, the population characteristics corresponding to the advertising plan, the object characteristics of the advertising object corresponding to the advertising plan, and the material characteristics of the advertising material corresponding to the advertising plan are used as at least one advertising feature corresponding to the advertising plan.

[0019] Optionally, the revenue estimation model is trained in the following manner:

[0020] Obtaining historical advertising plans and historical budgets and historical revenue values ​​corresponding to the historical advertising plans;

[0021] using the historical advertising plan as a sample advertising plan, the historical budget as a sample budget, and the historical revenue value as a sample revenue value;

[0022] determining at least one sample advertising characteristic of the sample advertising campaign;

[0023] At least one sample advertisement feature of the sample advertisement plan and the sample budget are used as input data, and the sample revenue value is used as a training label to train a revenue estimation model.

[0024] Optionally, it also includes:

[0025] Determine budget candidate sets corresponding to the multiple advertising plans respectively;

[0026] The at least one estimated revenue value generated when any advertising plan is allocated any candidate budget includes:

[0027] At least one estimated revenue value generated when any advertisement plan allocates any candidate budget in the corresponding preset candidate set is estimated.

[0028] Optionally, determining the budget candidate sets corresponding to the multiple advertising plans respectively includes:

[0029] For any advertising plan, according to the actual budget consumed by the advertising plan in the last delivery cycle, within the budget range corresponding to the actual budget, multiple candidate budgets are generated to form the budget candidate set;

[0030] The candidate budgets respectively matching the multiple advertising plans as the target budgets respectively corresponding to the multiple advertising plans include:

[0031] The candidate budgets respectively matched by the multiple advertising plans are used as the target budgets respectively corresponding to the multiple advertising plans in the current delivery cycle;

[0032] The method further comprises:

[0033] The target budgets respectively corresponding to the multiple advertising plans in the current delivery cycle are updated to the advertising delivery system, so that the advertising delivery system can perform promotion operations for the multiple advertising plans in the current delivery cycle according to their respective corresponding target budgets.

[0034] Optionally, taking at least one estimated revenue value corresponding to each of the multiple advertising plans satisfying a revenue value requirement as an optimization goal, determining candidate budgets that match each of the multiple advertising plans includes:

[0035] The candidate budgets that match the multiple advertising plans are determined by taking the maximization of the total value corresponding to the first estimated revenue values ​​respectively generated by the multiple advertising plans as the optimization goal, at least one second estimated revenue value respectively generated by the multiple advertising plans being within the revenue limit range, and the total budget of the candidate budgets respectively corresponding to the multiple advertising plans being within the budget limit range as the constraint conditions.

[0036] Optionally, it also includes:

[0037] updating the target budgets respectively corresponding to the multiple advertising plans to the advertising delivery system, so that the advertising delivery system performs promotion operations for the multiple advertising plans according to the target budgets respectively corresponding to the multiple advertising plans;

[0038] Obtain at least one revenue value generated by any advertisement plan in the advertisement delivery system within a predetermined time;

[0039] detecting whether the at least one benefit value satisfies a first benefit assessment requirement;

[0040] If not, the actual budget of the previous delivery cycle of the advertising plan is used as the target budget of the current delivery cycle and updated to the advertising delivery system.

[0041] Optionally, it also includes:

[0042] Delivering the multiple advertising plans according to the target budget in the simulation system, and obtaining at least one simulation revenue value generated by the multiple advertising plans;

[0043] determining whether the at least one simulation benefit value satisfies a second benefit assessment requirement;

[0044] If so, delivering the plurality of advertisement plans to an advertisement delivery system;

[0045] If not, output a warning message.

[0046] In a second aspect, an embodiment of the present application provides a computing device, including a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement the budget allocation method as described in the first aspect above.

[0047] In a third aspect, an embodiment of the present application provides a computer storage medium storing a computer program, and when the computer program is executed by a computer, the budget allocation method as described in the first aspect above is implemented.

[0048] In a fourth aspect, an embodiment of the present application provides a computer program product, including a computer program / instructions, which, when executed by a computer, implements the budget allocation method described in the first aspect above.

[0049] The embodiment of the present application is directed to multiple advertising plans of an advertiser; at least one estimated revenue value generated when any candidate budget is allocated to any advertising plan is estimated; the candidate budgets that match the multiple advertising plans are determined with the optimization goal of satisfying the revenue value requirement for at least one estimated revenue value corresponding to each of the multiple advertising plans; the candidate budgets that match the multiple advertising plans are used as the target budgets that correspond to the multiple advertising plans. In the embodiment of the present application, the estimated revenue value and the set revenue value requirements are used to perform optimization and solve, so that the target budgets that correspond to the multiple advertising plans can be determined. Under the constraint of the target budget, after the multiple advertising plans are launched, the revenue value requirements can be met with a high probability, and the revenue value requirements can be pre-set in combination with actual needs, thereby ensuring the rationality of budget allocation, ensuring the advertising delivery effect, and improving the flexibility and efficiency of advertising delivery.

[0050] These and other aspects of the present application will become more clearly understood in the description of the following embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0052] Figure 1 A flowchart of an embodiment of a budget allocation method provided by the present application is shown;

[0053] Figure 2 A schematic diagram of the system architecture of an advertisement processing system provided by the present application is shown;

[0054] Figure 3 A conceptual schematic diagram showing an advertisement processing process in a practical application of an embodiment of the present application is shown;

[0055] Figure 4 A schematic diagram showing the structure of an embodiment of a budget allocation device provided by the present application is shown;

[0056] Figure 5 A schematic diagram of the structure of an embodiment of a computing device provided by the present application is shown. DETAILED DESCRIPTION

[0057] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application.

[0058] In some of the processes described in the specification and claims of this application and the above-mentioned figures, multiple operations that appear in a specific order are included, but it should be clearly understood that these operations may not be executed in the order in which they appear in this article or executed in parallel. The serial numbers of the operations, such as 101, 102, etc., are only used to distinguish between different operations, and the serial numbers themselves do not represent any execution order. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions of "first", "second", etc. in this article are used to distinguish different messages, devices, modules, etc., do not represent the order of precedence, and do not limit the "first" and "second" to be different types.

[0059] The technical solution of the embodiment of the present application can be applied to the scenario of promoting an object through external advertising, which is the promotion behavior of placing the advertising plan for promoting the object to be promoted (hereinafter described as the advertising object) to an external advertising delivery system such as various media platforms, etc., and paying the relevant fees. Among them, the advertising object can be, for example, an object provided by an online system, such as a product provided by an e-commerce platform, etc., and of course, it can also be other offline or online objects that need to be promoted.

[0060] At present, advertisers usually deliver advertisements to the advertising delivery system in units of advertising plans. An advertising plan may be composed of one or more advertising contents generated by materials such as advertising objects, target users, and advertising materials. Advertising materials may refer to some recommendation information of the advertising object, which may be multimedia data such as video, text or image. The target population may refer to the user group to be promoted, for example, it may be a group of users in a specified age range, or a specified occupation, or a specified user, etc.

[0061] After allocating a budget to an advertising plan, the advertising delivery system can perform promotion operations based on the advertising plan under the budget constraint. If the cost of a certain advertising plan exceeds the budget or exceeds the cost range determined based on the budget, the promotion operation corresponding to the advertising plan can be suspended. In the traditional way, advertisers mostly set a fixed budget for each advertising plan, resulting in that the advertising plans with better performance cannot continue to be displayed in the delivery cycle due to budget constraints, while the advertising plans with poor performance may waste the budget, which ultimately leads to low efficiency of advertising delivery and affects the effect of advertising delivery.

[0062] In order to ensure the rationality of budget allocation and improve the effect of advertising, the inventor has proposed the technical solution of the present application after a series of studies. In the embodiment of the present application, for multiple advertising plans of advertisers; estimate at least one estimated revenue value generated when any candidate budget is allocated to any advertising plan; take at least one estimated revenue value corresponding to each of the multiple advertising plans to meet the revenue value requirement as the optimization goal, and determine the candidate budgets that match the multiple advertising plans respectively; use the candidate budgets that match the multiple advertising plans respectively as the target budgets corresponding to the multiple advertising plans respectively. In the embodiment of the present application, the estimated revenue value and the set revenue value requirements are used to perform optimization and solve, so that the target budgets corresponding to the multiple advertising plans can be determined. Under the constraint of the target budget, after the multiple advertising plans are delivered, the revenue value requirements can be met with a high probability, and the revenue value requirements can be pre-set in combination with actual needs, thereby ensuring the rationality of budget allocation, ensuring the effect of advertising, and improving the flexibility and efficiency of advertising.

[0063] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of this application.

[0064] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0065] It should be noted that the technical solution of the embodiment of the present application is applicable to a network virtual environment. The users described generally refer to "virtual users". Real users can register a user account in the server through registration to obtain a user identity in the network environment.

[0066] The implementation details of the technical solution of the embodiment of the present application are elaborated in detail below.

[0067] Figure 1 This is a flowchart of an embodiment of a budget allocation method provided by the present application. The method may include the following steps:

[0068] 101: Decide on multiple advertising plans.

[0069] An advertising plan is mainly composed of one or more advertising contents generated by advertising objects, target users, and advertising materials. Advertisers can deliver advertising plans to external advertising delivery systems such as various media platforms.

[0070] The multiple advertising plans may be all advertising plans generated by the advertiser, so that budget allocation for a single advertising plan can be achieved from an overall perspective.

[0071] 102: Estimate at least one estimated revenue value generated when any candidate budget is allocated to any advertising plan.

[0072] The revenue value may refer to, for example, click-through rate, conversion rate, exposure, daily active users, GMV (Gross Merchandise Volume) or ROI (Return on Investment), etc. At least one estimated revenue value may include one or more indicators such as click-through rate, conversion rate, exposure, daily active users, GMV and ROI.

[0073] Click-through rate may refer to the ratio of the number of clicks corresponding to the advertising plan to the number of displays per unit time when the promotion operation based on the advertising plan is to display advertising information. Conversion rate may refer to the ratio of the number of interactive behaviors such as the number of purchases to the number of displays for the advertising object corresponding to the advertising plan per unit time. Exposure may refer to the number of displays of the advertising plan per unit time. Daily active users may refer to the daily new users attracted by the advertising plan. GMV may refer to the total transaction amount of the goods promoted by the advertising plan per unit time when the advertising object is a commodity; ROI may refer to the ratio of GMV to the investment cost.

[0074] 103: Taking at least one estimated revenue value corresponding to each of the multiple advertising plans as satisfying a revenue value requirement as an optimization goal, determine candidate budgets that match each of the multiple advertising plans.

[0075] 104: Using the candidate budgets respectively matched with the multiple advertising plans as the target budgets respectively corresponding to the multiple advertising plans.

[0076] The target budget corresponding to each advertising plan can be updated to the advertising delivery system so that the advertising delivery system can perform promotion operations based on the advertising plan under the constraint of the target budget.

[0077] In the embodiment of the present application, the optimal solution is obtained by estimating the revenue value and the set revenue value requirements, so that the target budgets corresponding to multiple advertising plans can be determined. Under the target budget constraints, after the multiple advertising plans are launched, the revenue value requirements can be met with a high probability. The revenue value requirements can be pre-set based on actual needs, thereby ensuring the rationality of budget allocation, ensuring the effectiveness of advertising delivery, and improving the flexibility and efficiency of advertising delivery.

[0078] In some embodiments, estimating at least one estimated revenue value generated when any advertising plan is allocated to any candidate budget may include: using at least one revenue estimation model to estimate at least one estimated revenue value generated when any advertising plan is allocated to any candidate budget. Each revenue estimation model may be obtained by training based on sample advertising plans, sample budgets corresponding to the sample advertising plans, and sample revenue values.

[0079] Among them, each revenue estimation model can be used to estimate the estimated revenue value corresponding to a revenue type. The revenue type can refer to, for example, click-through rate type, conversion rate type, exposure type, daily active user type, GMV type or ROI type.

[0080] In some embodiments, using at least one revenue estimation model, estimating at least one estimated revenue value generated when any advertising plan is allocated to any candidate budget may include: determining at least one advertising feature corresponding to any advertising plan; inputting at least one advertising feature and any candidate budget into at least one revenue estimation model to obtain at least one estimated revenue value generated when the advertising plan is allocated to the candidate budget.

[0081] The embodiment of the present application can effectively realize the profit estimation by using the machine learning model. The above profit estimation model can be, for example, a classifier, a support vector machine or various neural network models or AI (artificial intelligence) models, etc., and the present application does not limit this.

[0082] In some embodiments, determining at least one advertising feature corresponding to any advertising plan may include: determining at least one average revenue value generated by any advertising plan within a historical time period; using one or more of the at least one average revenue value, population characteristics corresponding to the advertising plan, object characteristics of the advertising object corresponding to the advertising plan, and material characteristics of the advertising material corresponding to the advertising plan as at least one advertising feature corresponding to the advertising plan.

[0083] For example, the revenue value may be click-through rate, conversion rate, exposure, daily active users, GMV or ROI, etc., and the at least one average revenue value generated by the advertising plan in the historical time period may include one or more of the daily average click-through rate, average conversion rate, average exposure, average daily active users, average GMV and average ROI. The population characteristics corresponding to the advertising plan may refer to the characteristics of the population to which the advertising plan is delivered, such as the number, gender, age, region and preference of the population. The advertising object corresponding to the advertising plan may refer to a commodity, and the object characteristics of the advertising object may include object attributes such as object category, etc. The material characteristics of the advertising material corresponding to the advertising plan may refer to, for example, the material form such as video, image and text, the material content type and the material design style, etc.

[0084] In some embodiments, the revenue estimation model can be trained in the following manner: obtain historical advertising plans and the historical budgets and historical revenue values ​​corresponding to the historical advertising plans; use the historical advertising plans as sample advertising plans, the historical budgets as sample budgets, and the historical revenue values ​​as sample revenue values; determine at least one sample advertising feature of the sample advertising plan; use at least one sample advertising feature of the sample advertising plan and the sample budget as input data, and the sample revenue values ​​as training labels to train the revenue estimation model.

[0085] The embodiment of the present application trains the revenue estimation model by combining historical data, so that the revenue estimation model can learn from the historical data and understand the patterns and trends in the data, and can make accurate predictions about future or unknown data based on this, thereby improving the accuracy of the estimation.

[0086] In some embodiments, the method may also include: obtaining a first test sample; the first test sample includes a test advertising plan and a test revenue value corresponding to the test advertising plan; using the revenue estimation model, estimating the estimated revenue value corresponding to the test advertising plan; determining whether the revenue estimation model meets the model requirements based on the estimated revenue value and the test revenue value; if not, adjusting the revenue estimation model.

[0087] If the revenue estimation model meets the model requirements, the revenue estimation model can be used to estimate the estimated revenue value generated by any advertising plan.

[0088] The first test sample may also come from a historical delivery record, the test advertising plan may also be a historical advertising plan, and the test revenue value may also be a historical revenue value generated by a historical advertising plan.

[0089] The model requirement may be, for example, that the difference between the estimated revenue value and the test revenue value is within an allowable range, etc. If the revenue estimation model does not meet the model requirement, the revenue estimation model may be adjusted, for example, the step of training the value estimation model using the sample advertising plan and the sample revenue value may be returned to continue the execution to retrain the revenue estimation model or re-adjust the training sample, etc.

[0090] In addition, first adjustment prompt information may be output to prompt the user whether to adjust the revenue estimation model, etc. The first adjustment prompt information may include information on the difference between the estimated revenue value and the test revenue value, etc. The first adjustment prompt information may be sent to the user terminal so that the user terminal displays the first adjustment prompt information, etc. in the interactive interface.

[0091] In the embodiment of the present application, by using the first test sample to test the revenue estimation model, the performance and generalization ability of the model can be evaluated to ensure the accuracy of the model.

[0092] In some embodiments, the method may further include: determining budget candidate sets corresponding to a plurality of advertising plans respectively.

[0093] Estimating at least one estimated revenue value generated when any advertisement plan is allocated with any candidate budget may include: estimating at least one estimated revenue value generated when any advertisement plan is allocated with any candidate budget in its corresponding budget candidate set.

[0094] The budget candidate set may be generated from a plurality of pre-set candidate budgets or generated based on historical budgets.

[0095] In some embodiments, the budget allocation of the advertising plan can be performed according to the delivery cycle, for example, a delivery cycle can be one day, one week, one month, etc. The above-mentioned determination of the budget candidate sets corresponding to the multiple advertising plans can include: for any advertising plan, according to the actual budget consumed by the advertising plan in the previous delivery cycle, within the budget range corresponding to the actual budget, generating multiple candidate budgets to form a budget candidate set.

[0096] Using the candidate budgets matched to the multiple advertising plans as the target budgets corresponding to the multiple advertising plans may include: using the candidate budgets matched to the multiple advertising plans as the target budgets corresponding to the multiple advertising plans in the current delivery cycle.

[0097] The method may also include: updating the target budgets corresponding to the multiple advertising plans in the current delivery cycle to the advertising delivery system, so that the advertising delivery system can display advertisements for the multiple advertising plans in accordance with their corresponding target budgets in the current delivery cycle.

[0098] Among them, for any advertising plan, according to the actual budget consumed by the advertising plan in the previous delivery cycle, within the budget range corresponding to the actual budget, multiple candidate budgets are generated to form a budget candidate set. The budget range can be set first, and the actual budget consumed by the advertising plan in the previous delivery cycle is taken as a candidate budget. The actual budget is increased or decreased to obtain other candidate budgets until the boundary of the budget range is reached, and all candidate budgets constitute the budget candidate set.

[0099] For example, a delivery cycle can be one day, and the budget corresponding to advertising plan i on day t can be selected from the budget candidate set S t,i [S t-1,i -d,S t-1,i -(d-1), ···, S t-1,i -1, S t-1,i , S t-1,i +1,···,S t-1,i +d], where S t-1,i is the actual budget consumed by advertising plan i on day t-1, and the budget candidate set S t,i The maximum candidate budget S t-1,i +d is the actual budget consumed by advertising plan i on day t-1 plus d, and the minimum candidate budget S t-1,i -d is the actual budget consumed by advertising plan i on day t-1 minus d.

[0100] In some embodiments, the optimization goal is that at least one estimated revenue value corresponding to each of the multiple advertising plans meets the revenue value requirement, and determining the candidate budgets that match each of the multiple advertising plans may include: taking the total value corresponding to the first estimated revenue values ​​respectively generated by the multiple advertising plans as the optimization goal to be maximal, at least one second estimated revenue value respectively generated by the multiple advertising plans being within a revenue limit range, and the total budget of the candidate budgets corresponding to the multiple advertising plans being within a budget limit range as constraints, to determine the candidate budgets that match each of the multiple advertising plans.

[0101] The total value corresponding to the first estimated revenue values ​​generated by the multiple advertising plans refers to the overall revenue value corresponding to the advertiser, which is obtained by adding the first estimated revenue values ​​of the multiple advertising plans.

[0102] The total budget of the candidate budgets corresponding to the multiple advertising plans refers to the overall budget corresponding to the advertiser, which is obtained by adding up the candidate budgets of all the advertising plans.

[0103] The total budget of the candidate budgets respectively corresponding to the multiple advertising plans being within the budget limit range may mean that the total budget of the candidate budgets respectively corresponding to the multiple advertising plans does not exceed a preset total budget.

[0104] Among them, the first estimated revenue value may be an estimated revenue value of a specified revenue type, such as ROI, etc., and the second estimated revenue value may be an estimated revenue value of other specified revenue types different from the first estimated revenue value, such as click-through rate, conversion rate, daily active users, exposure or GMV, etc. Among them, the second estimated revenue value may include one or more. In the case of multiple second estimated revenue values, the constraint condition may be that each second estimated revenue value is within the respective corresponding predetermined revenue range. For example, ROI can be used as the first estimated revenue value, and the total value corresponding to the first estimated revenue values ​​generated by multiple advertising plans is the total ROI of multiple advertising plans. One or more estimated revenue values ​​of click-through rate, conversion rate, exposure, daily active users and GMV can be used as at least one second estimated revenue value.

[0105] Taking click-through rate, conversion rate, exposure, daily active users and GMV as the second estimated revenue value as an example, at least one second estimated revenue value generated by multiple advertising plans respectively being within the revenue limit range can mean that the total value of each second estimated revenue value corresponding to the multiple advertising plans is within the revenue limit range, for example, the total click-through rate reaches the second predetermined click-through rate, the total conversion rate reaches the second predetermined conversion rate, the total daily active users reach the second predetermined daily active users, the total exposure reaches the second predetermined exposure, and the total GMV reaches the second predetermined GMV. The total value of each second estimated revenue value can be obtained by adding the same second estimated revenue value generated by multiple advertising plans respectively.

[0106] Among them, the total exposure generated by multiple advertising plans can be obtained by adding up the exposure of each advertising plan. The total click-through rate generated by multiple advertising plans can be obtained by calculating the ratio of the total number of clicks of multiple advertising plans to the total exposure in a unit time. The total conversion rate generated by multiple advertising plans can be obtained by calculating the ratio of the total number of interactive behaviors such as the number of purchases to the advertising objects corresponding to the multiple advertising plans in a unit time to the total exposure. The total daily active users generated by multiple advertising plans can be obtained by adding up the daily new users attracted by each advertising plan. The total GMV generated by multiple advertising plans can be obtained by adding up the GMV generated by each advertising plan. The total ROI generated by multiple advertising plans can be obtained by calculating the ratio of the total GMV generated by multiple advertising plans to the total budget of multiple advertising plans.

[0107] For example, suppose S t,i represents the candidate budget set corresponding to advertising plan i on day t, S t,ij Denotes the candidate budget set S corresponding to advertising plan i on day t t,i In addition, the budget selection matrix A can be established. ij =1 indicates that advertising plan i is in the candidate budget set S t,i Select the jth candidate budget, A ij =0 means that advertising plan i is in the candidate budget set S t,i The jth candidate budget is not selected. Establish the first revenue estimation matrix R, R ij Indicates the first estimated revenue value corresponding to the jth candidate budget selected by advertising plan i. Establish a second estimated revenue matrix O, for example, there may be k second estimated revenue values It represents the first and second estimated revenue value corresponding to the jth candidate budget selected by advertising plan i. It represents the kth second estimated revenue value corresponding to the jth candidate budget selected by advertising plan i. Then the target budgets matching multiple advertising plans can be obtained by solving the following objective function:

[0108] max∑ i ∑ j R ij ·A ij

[0109]

[0110] ∑ i ∑ j S t,ij ·A ij ≤Total Budget

[0111]

[0112] ···

[0113]

[0114] where ∑ j A ij =1, indicating that each advertising plan will select a candidate budget from the corresponding candidate budget set. The total value corresponding to the first estimated revenue value generated by multiple advertising plans ∑ i ∑ j R ij ·A ij The maximum is the optimization target, and the total value corresponding to the kth second estimated revenue value generated by multiple advertising plans Located in the kth profit limit range, ∑ i ∑ j S t,ij ·A ij The constraint condition of not exceeding the predetermined total budget is used to solve the problem to determine candidate budgets that match multiple advertising plans. For example, the solution may be solved by using a branch and cut method, which is not limited in the present application.

[0115] In some embodiments, the method may also include: obtaining at least one revenue value generated by any advertising plan within a predetermined time in the advertising delivery system; detecting whether at least one revenue value meets a first revenue evaluation requirement; if not, using the actual budget of the advertising plan in the previous delivery cycle as the target budget for the current delivery cycle, and updating it to the advertising delivery system.

[0116] The above-mentioned at least one revenue value may include one or more of click-through rate, conversion rate, exposure, daily active users, GMV and ROI. The first revenue estimation requirement may, for example, include the revenue conditions corresponding to each revenue requirement, for example, may include: whether the click-through rate reaches a first predetermined click-through rate, whether the conversion rate reaches a first predetermined conversion rate, whether the exposure reaches a first predetermined exposure, whether the daily active users reach a first predetermined daily active users, whether the GMV reaches a first predetermined GMV, and whether the ROI reaches a first predetermined ROI, etc. One or more of the above.

[0117] Therefore, detecting whether at least one revenue value meets the revenue evaluation requirements may include detecting whether the click-through rate reaches a first predetermined click-through rate, whether the conversion rate reaches a first predetermined conversion rate, whether the exposure reaches a first predetermined exposure, whether the daily active users reach a first predetermined daily active users, whether the GMV reaches a first predetermined GMV, and whether the ROI reaches a first predetermined ROI, etc.

[0118] If any of the at least one revenue value does not meet the corresponding revenue condition, for example, the actual budget S consumed by advertising plan i on day t-1 can be t-1,i As the budget corresponding to advertising plan i on day t.

[0119] In order to facilitate the management and control of advertising plans, the present application embodiment also provides an advertising processing system, such as Figure 2 In the system architecture diagram shown, the advertisement processing system may include a user terminal 201 and a server terminal 202. A connection may be established between the user terminal 201 and the server terminal 202 via a network. The network provides a medium for a communication link between the user terminal 201 and the server terminal 202. The network may include various connection types, such as wired or wireless communication links or optical fiber cables, etc. The user terminal 201 may interact with the server terminal 202 via the network to receive or send messages, etc. The server terminal 202 may be connected to the advertisement delivery system 200 so that the advertisement plan may be directly delivered to the advertisement delivery system 200 based on the connection.

[0120] Among them, the user terminal 201 can be a browser, an APP (Application), or a web application such as an H5 (HyperText Markup Language 5, Hypertext Markup Language Version 5) application, or a light application (also known as a mini-program, a lightweight application) or a cloud application, etc. The user terminal 201 can be deployed in an electronic device and needs to rely on the device to run or some apps in the device to run, etc. The electronic device can, for example, have a display screen and support information browsing, such as a personal mobile terminal such as a mobile phone, a tablet computer, a personal computer, a desktop computer, a smart speaker, a smart watch, etc. For ease of understanding, Figure 2The user end is mainly represented by a device image. Various other types of applications can usually be configured in electronic devices, such as human-computer dialogue applications, model training applications, text processing applications, web browser applications, shopping applications, search applications, instant messaging tools, email clients, social platform software, etc. Electronic devices can refer to devices used by users, which have the functions of computing, surfing the Internet, communicating, etc. required by users, such as mobile phones, tablet computers, personal computers, wearable devices, etc. Electronic devices can usually include at least one processing component and at least one storage component. Electronic devices may also include basic configurations such as network card chips, IO buses, audio and video components, which are not limited in this application. Optionally, according to the implementation form of the electronic device, some peripheral devices may also be included, such as keyboards, mice, input pens, printers, etc., which are not limited in this application.

[0121] The server 202 may include servers that provide various services, such as a server that provides support for model training, or a server that processes information sent by a user.

[0122] It should be noted that the server 202 can be implemented as a distributed server cluster consisting of multiple servers, or as a single server. The server can also be a server of a distributed system, or a server combined with a blockchain. The server can also be a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (CDN), and big data and artificial intelligence platforms, or an intelligent cloud computing server or intelligent cloud host with artificial intelligence technology.

[0123] Figure 1 The budget allocation method provided in the illustrated embodiment may be executed by the server 202. The advertising plan may be provided by the user through the user terminal 201, etc. Therefore, in some embodiments, the method may further include:

[0124] Providing an interactive interface to the user terminal; determining multiple advertising plans provided by the user according to the interactive information sent by the user terminal.

[0125] In addition, the user can also specify the revenue type through the interactive interface to facilitate the determination of estimated targets, etc. Therefore, in some embodiments, multiple advertising plans and at least one revenue type sent by the user end can be obtained; the at least one revenue type is provided by the user in the interactive interface.

[0126] The estimating at least one estimated revenue value generated when any advertising plan is allocated to any candidate budget may include: estimating at least one estimated revenue value generated when any advertising plan is allocated to any candidate budget based on at least one revenue type.

[0127] The at least one revenue type may include, for example, one or more of a ROI type, a GMV type, a daily active user type, an exposure type, a conversion rate type, and a click-through rate type.

[0128] In some possible implementations, the server may send advertisement selection prompt information to the user so that the user displays the advertisement selection prompt information in the interactive interface. The advertisement selection prompt information may include an input control for sensing multiple advertisement plans set by the user, or the advertisement selection prompt information may include an advertisement plan list and a selection control for sensing multiple advertisement plans selected by the user in the advertisement plan list. Of course, the implementation scheme of the present application is not limited thereto.

[0129] In addition, the server may send revenue type prompt information to the user so that the user can display the revenue type prompt information in the interactive interface. The revenue type prompt information may include an input control for sensing at least one revenue type input by the user, or the revenue type prompt information may include multiple revenue types and a selection control for sensing at least one revenue type selected by the user from multiple revenue types. Of course, the implementation scheme of the present application is not limited to this.

[0130] In order to ensure the accuracy of the estimated results and to guarantee the final delivery effect, in some embodiments, such as Figure 2 As shown in , the advertisement processing system may further include a simulation system 203, and the method may further include:

[0131] In the simulation system, multiple advertising plans are delivered according to the target budget, and at least one simulated revenue value generated by the multiple advertising plans is obtained; it is determined whether the at least one simulated revenue value meets the second revenue evaluation requirement; if so, the multiple advertising plans are delivered to the advertising delivery system; if not, a warning prompt message is output.

[0132] The simulation system is a simulation model of an advertisement delivery system, which is used to simulate the delivery of multiple advertisement plans to be delivered, and to determine whether to actually deliver the advertisements to the advertisement delivery system in combination with the simulation revenue value generated in the simulation system.

[0133] The second revenue evaluation requirement can be set in combination with actual needs, etc., to evaluate each simulated revenue value of each advertising plan or the comprehensive value of at least one simulated revenue value corresponding to each advertising plan or the total value of each simulated revenue value of multiple advertising plans. The comprehensive value can be obtained by weighted processing, and the total value can be obtained by summing or weighted summing, etc.; for example, when at least one simulated revenue value includes ROI, GMV or conversion rate, the second revenue evaluation requirement can be, for example, that the ROI, GMV or conversion rate of each advertising plan reaches a specified evaluation threshold, etc., or that the total ROI, total GMV or total conversion rate corresponding to multiple advertising plans reaches a specified evaluation threshold, or that the comprehensive value of each advertising plan reaches a specified evaluation threshold, etc. Of course, it can also be set based on the revenue value requirement, etc., and this application does not limit this.

[0134] At least one simulated benefit value may be of the same benefit type as at least one estimated benefit value. Of course, benefit values ​​of different benefit types may also be used for verification and evaluation.

[0135] If each advertising plan meets the second revenue evaluation requirement, multiple advertising plans can be delivered to the advertising delivery system. Otherwise, if any advertising plan does not meet the second revenue evaluation requirement, a warning prompt message can be output.

[0136] Optionally, the outputting of the warning prompt information may be sending the warning prompt information to the user terminal, so that the user terminal can display the warning prompt information in the interactive interface. The warning prompt information can be used to notify the user that multiple advertising plans have failed to be verified, etc., and can also include at least one simulated revenue value corresponding to each advertising plan, so that the user can analyze the reasons and readjust the advertising plan and / or revenue type, etc.

[0137] In addition, the inventors found in their research that in actual applications, advertisers usually generate advertising plans based on given materials such as advertising objects, advertising materials and / or target groups. In traditional methods, advertisers usually generate advertising plans based on given materials in combination with manual experience. The advertising plan may include multiple advertising contents, and the advertising delivery system outputs the corresponding advertising contents according to its own logic to achieve promotion operations. However, the manual generation method may not be coupled with the external advertising delivery system and may not be consistent with actual needs. Therefore, how to generate a reasonable advertising plan to improve the delivery effect has also become a technical problem that needs to be solved at present.

[0138] In order to improve the rationality of generating an advertising plan, in some embodiments, the method may further include:

[0139] Determine a material set; estimate at least one estimated advertising value generated by an advertising plan generated based on any material in the material set; find multiple target materials so that at least one estimated advertising value respectively generated by multiple advertising plans generated based on the multiple target materials meets the advertising value requirement; generate multiple advertising plans using the multiple target materials.

[0140] Among them, the materials in the material set can be provided by users, that is, advertisers. Users can be providers of advertising objects, such as merchants. Of course, they can also be operators of online systems that provide object exchange behaviors, such as e-commerce platform operators, etc. This application does not limit this.

[0141] The user can provide the materials in the material set by means of code configuration. In addition, for the convenience of operation, an interactive interface can also be provided to provide the materials in the material set by means of interface input, etc., which will be described in detail in the following embodiments.

[0142] The material in a material set may refer to a single element such as an advertising object, advertising material, or a target population, or it may be a combined element composed of multiple elements of an advertising object, advertising material, and a target population. The combined element may be composed of any advertising object, any advertising material, and / or any target population, etc. This application does not limit this.

[0143] One material can generate one advertising plan. The material can include one or more elements of advertising objects, advertising materials, and target groups, and one element can belong to one or more advertising plans.

[0144] An advertising plan may correspond to one or more materials.

[0145] In the embodiment of the present application, based on the advertising plan generated by the material, the advertising value that may be generated can be estimated. For the convenience of description, the estimated result is named "estimated advertising value".

[0146] In an embodiment of the present application, the advertising value may refer to, for example, click-through rate, conversion rate, exposure, daily active users, GMV or ROI, and at least one estimated advertising value may include one or more indicators such as click-through rate, conversion rate, exposure, daily active users, GMV and ROI.

[0147] The advertising value may be the same as or different from the above-mentioned revenue value.

[0148] Optionally, after a plurality of advertising plans are generated based on a plurality of target materials, the plurality of advertising plans may be delivered to an advertising delivery system so that the advertising delivery system can perform corresponding promotion operations based on each advertising plan.

[0149] Among them, each target material can belong to an advertising plan. Combined with the above description, it can be known that the target material can include a single element or a combination of multiple elements in the advertising object, target population, and advertising material. When the target material includes some elements, the advertising plan can be generated using the target material and the remaining elements except the target material, and the remaining elements except the target material can be pre-set, etc. Of course, in a specific implementation, the target material can include all the elements constituting the advertising plan, such as the advertising object, target population, and the combination elements of the advertising material, and this application does not limit this.

[0150] In the embodiment of the present application, the advertising value requirement can be used as the optimization target, so that multiple target materials that meet the optimization target can be found, and then an advertising plan can be generated based on each target material, thereby obtaining multiple advertising plans. By estimating the advertising value and the set advertising value requirements, the optimization solution is performed, so that after the multiple advertising plans generated based on the found multiple target materials are put into use, the advertising value requirements can be met with a high probability, and the advertising value requirements can be pre-set in combination with the actual advertising generation requirements or the delivery effect requirements, thereby ensuring the rationality of the advertising plan generation.

[0151] In some embodiments, the at least one estimated advertising value generated by estimating the advertising plan generated based on any one of the material sets may include:

[0152] At least one value estimation model is used to estimate at least one estimated advertising value generated by an advertising plan generated based on any material in the material set.

[0153] The value estimation model may be obtained by training based on sample materials and sample advertisement values ​​corresponding to the sample materials.

[0154] Each value estimation model can be used to estimate the estimated advertising value corresponding to a value type. The value type can refer to, for example, click-through rate type, conversion rate type, exposure type, daily active user type, GMV type, or ROI type.

[0155] The embodiment of the present application can effectively realize value estimation by using a machine learning model. The above value estimation model can be, for example, a classifier, a support vector machine, or various neural network models or AI (artificial intelligence) models, etc., which is not limited in the present application.

[0156] In some embodiments, the value estimation model can be pre-trained in the following manner:

[0157] Determine the historical advertising plan and the historical advertising value generated by the historical advertising plan; use the materials involved in the historical advertising plan as sample materials, and the historical advertising value as sample advertising value; use the sample materials and sample advertising value to train the value estimation model.

[0158] The embodiment of the present application trains the value estimation model by combining historical data, so that the value estimation model can learn from the historical data and understand the patterns and trends in the data, and can make accurate predictions about future or unknown data based on this, thereby improving the accuracy of the estimation.

[0159] Combined with the above description, it can be known that each material can only belong to one advertising plan. The at least one estimated advertising value generated by the advertising plan generated by each material can be considered as the at least one estimated advertising value corresponding to the material. In actual applications, since the link from the material to the advertising revenue is long, in order to ensure accuracy, the at least one estimated advertising value can include multiple estimated advertising values. The multiple estimated advertising values ​​can include the estimated advertising values ​​corresponding to different stages in the link from the material to the advertising revenue. For example, the multiple estimated advertising values ​​can include click-through rate, conversion rate and ROI. For the convenience of processing, the multiple estimated advertising values ​​corresponding to each material can be weighted, and the weighted processing result is used as the material value. Therefore, in some embodiments, the method can also include:

[0160] At least one estimated advertising value corresponding to any material is weighted to generate a material value corresponding to the material.

[0161] Among them, the weighted processing method may include, for example, weighted sum, weighted average or weighted power calculation, etc. The weighted sum may be the sum of the products obtained by multiplying each estimated advertising value by its weight; the weighted average may refer to the sum of the products obtained by multiplying each estimated advertising value by its weight divided by the sum of all weights; the weighted power calculation may refer to taking the weight of each estimated advertising value as a power to perform power calculation on each estimated advertising value, and then multiplying the power calculations of each estimated advertising value, etc. Of course, the present application is not limited to this.

[0162] After obtaining the material value, in one implementation method, the above-mentioned searching for multiple target materials so that at least one estimated advertising value respectively generated by multiple advertising plans generated based on the multiple target materials meets the advertising value requirement can be: searching for multiple target materials so that the material values ​​corresponding to the multiple target materials respectively meet the advertising value requirement.

[0163] The advertising value requirement may be, for example, that the sum of the material values ​​is maximized, so as to ensure that multiple advertising plans generated based on multiple target materials can bring better returns, so as to ensure the delivery effect.

[0164] As another implementation method, since an advertising plan may involve one or more materials, in actual applications, if multiple materials are involved, the multiple materials will be considered homogeneous by the advertising delivery system, which will cause certain losses. Therefore, it is necessary to ensure the homogeneity of materials between the same advertising plan. The above-mentioned search for multiple target materials so that at least one estimated advertising value generated by multiple advertising plans generated based on multiple target materials meets the advertising value requirement can be: taking the material value difference between materials in the same advertising plan to meet the difference requirement as the optimization goal, and finding at least two target materials that match multiple advertising plans respectively. This implementation method can ensure the homogeneity of at least two target materials in multiple advertising plans.

[0165] The material value differences between materials can be expressed by calculating the degree of dispersion between material values, and the degree of dispersion can be expressed by data indicators such as variance, standard deviation or mean deviation, etc. Under the condition of ensuring the homogeneity of materials in the same advertising plan generated, the difference requirement can refer to the minimum degree of dispersion, such as the minimum variance, etc.

[0166] In addition, as another implementation method, the materials in the material set can be grouped in different division methods, and a corresponding advertising plan can be generated by using the materials in the same group as the target materials. The advertising plan corresponding to each group can be considered as an advertising plan that meets the advertising value requirements.

[0167] There are many ways of grouping, such as equal frequency grouping, that is, grouping multiple materials whose values ​​are different within a specified difference range into one group, so as to ensure the homogeneity of the materials in each group and the final delivery effect. Of course, according to different actual needs, grouping can also be performed by using equal distance grouping or clustering grouping such as K-means (a clustering algorithm), which is not limited in this application.

[0168] In addition, to ensure delivery efficiency and delivery effect, the advertising value requirement may include optimization objectives and constraints. In some embodiments, the above-mentioned searching for multiple target materials so that at least one estimated advertising value generated by multiple advertising plans generated based on the multiple target materials respectively meets the advertising value requirement may include:

[0169] With the first estimated advertising values ​​respectively generated by multiple advertising plans satisfying the advertising value requirements as the optimization goal, and the second estimated advertising values ​​respectively generated by multiple advertising plans or the sum of the second estimated advertising values ​​being within a predetermined value range as the constraint conditions, the target materials matching the multiple advertising plans are obtained.

[0170] The first estimated advertising value may be an estimated advertising value of a specified value type, such as ROI, etc. The second estimated advertising value may be an estimated advertising value of other specified value types different from the first estimated advertising value, such as click-through rate, conversion rate, exposure, daily active users or GMV, etc. Among them, the second estimated advertising value may include one or more. In the case of multiple second estimated advertising values, the constraint condition may be that each second estimated advertising value is within the respective corresponding predetermined value range. Of course, it may also be that the sum of the second estimated advertising values ​​generated by multiple advertising plans is within the predetermined value range. Among them, the sum of the second estimated advertising values ​​generated by multiple advertising plans is within the predetermined value range, which means that the overall advertising value corresponding to the advertiser is constrained from an overall perspective. The sum of the second estimated advertising values ​​generated by multiple advertising plans can be obtained by adding the second estimated advertising values ​​generated by multiple advertising plans.

[0171] The predetermined value range may be used to constrain the minimum and maximum values ​​of the second estimated advertising value or the total second estimated advertising value.

[0172] In combination with the previous description, it can be seen that the advertising value requirement can be, for example, the maximum sum of the first estimated advertising values ​​generated by multiple advertising plans, or the first estimated advertising value can be used as the material value corresponding to the material included in the advertising plan. The advertising value requirement can also be that the difference in material value between materials in the same advertising plan satisfies the difference requirement.

[0173] Among them, the sum of the first estimated advertising values ​​generated by multiple advertising plans is the largest, which means evaluating the overall advertising value corresponding to the advertiser from an overall perspective. The sum of the first estimated advertising values ​​generated by multiple advertising plans is obtained by adding the first estimated advertising values ​​generated by multiple advertising plans.

[0174] Optionally, the material value differences between materials in each advertising plan may be calculated first, and the advertising value requirement may be that the sum of the material value differences corresponding to multiple advertising plans satisfies the difference requirement. When the material value difference is a variance, the difference requirement may be, for example, that the sum of the variances corresponding to multiple advertising plans is the minimum.

[0175] In practical applications, for example, multiple target materials can be obtained by solving the following objective function:

[0176]

[0177] in, Represents the matching matrix between materials and advertising plans;

[0178] M represents the number of advertising plans; D represents a differentiation function such as a variance function used to solve the material value differences between materials in the same advertising plan, that is, variance. U represents the material value; L represents the material set; l represents the material; C[m:] represents the matching relationship between material l and the mth advertising plan, C[m:] = 0, indicating that material l has not generated the mth advertising plan and has no matching relationship; C[m:] = 1, indicating that material l has generated the mth advertising plan and has a matching relationship. It means that a material belongs to one advertising plan and has a matching relationship with only one advertising plan.

[0179] By solving the above objective function, the matching matrix can be obtained. According to the values ​​0 and 1 in the matching matrix, the target materials that match the M advertising plans can be determined.

[0180] Wherein, U represents the value of the material, and the value of the material can be obtained, for example, in the following manner:

[0181] Among them, p i represents the estimated value of advertising, α i Represents weight, i=1, 2,...k, k represents the estimated advertising value.

[0182] In addition, in combination with the corresponding description in the previous text, it can be seen that the material set can be provided by the user; and the advertising plan generated based on the target material can be delivered to the advertising delivery system.

[0183] Therefore, in some embodiments, the method may further include:

[0184] Provide an interactive interface to the user end so that the user can provide a material collection in the interactive interface; obtain the material collection sent by the user end.

[0185] In addition, the user can also specify the value type through the interactive interface to facilitate the determination of the estimated target, etc. Therefore, in some embodiments, the material set and at least one value type sent by the user end can be obtained; the at least one value type is provided by the user in the interactive interface;

[0186] The at least one estimated advertising value generated by the advertising plan generated based on any one of the material sets includes:

[0187] At least one estimated advertising value corresponding to at least one advertising value type generated by an advertising plan generated based on any one material in the material set is estimated.

[0188] In some possible implementations, the server may send material prompt information to the user so that the user can display the material prompt information in the interactive interface. The material prompt information may include an input control for sensing a material set input by the user, or the material prompt information may include a material list and a selection control for sensing a material set selected by the user in a specified material list. Of course, the implementation scheme of the present application is not limited thereto.

[0189] In addition, the server may send value type prompt information to the user so that the user can display the value type prompt information in the interactive interface. The value type prompt information may include an input control for sensing at least one value type input by the user, or the value type prompt information may include multiple value types and a selection control for sensing at least one value type selected by the user from multiple value types. Of course, the implementation scheme of the present application is not limited to this.

[0190] In order to ensure the accuracy of the estimated results and guarantee the final delivery effect, in some embodiments, the method may further include:

[0191] Multiple advertising plans are placed in the simulation system, and at least one simulated advertising value generated by each of the multiple advertising plans is obtained; whether the at least one simulated advertising value generated by each of the multiple advertising plans meets the first value evaluation requirement is determined; if so, the multiple advertising plans are placed in the advertising placement system; if not, a first prompt message is output.

[0192] The simulation system is a simulation model of an advertisement delivery system, which is used to simulate the delivery of multiple advertisement plans to be delivered, and to determine whether to actually deliver the advertisements to the advertisement delivery system based on the simulated advertisement values ​​generated in the simulation system.

[0193] The first value evaluation requirement can be set in combination with actual needs, etc., to evaluate each simulated advertising value of each advertising plan or the comprehensive value of at least one simulated advertising value corresponding to each advertising plan, or the total value of each simulated advertising value of multiple advertising plans. The comprehensive value can be obtained by weighted processing, and the total value can be obtained by summing or weighted summing, etc.; for example, when at least one simulated advertising value includes ROI, GMV or conversion rate, the first value evaluation requirement can be, for example, that the ROI, GMV or conversion rate of each advertising plan reaches a specified evaluation threshold, etc., or that the total ROI, total GMV or total conversion rate corresponding to multiple advertising plans reaches a specified evaluation threshold, or that the comprehensive value of each advertising plan reaches a specified evaluation threshold, etc. Of course, it can also be set based on advertising value requirements, etc., and this application does not limit this.

[0194] At least one simulated advertisement value may be of the same value type as at least one estimated advertisement value. Of course, advertisement values ​​of different value types may also be used for verification and evaluation.

[0195] If each advertising plan meets the first value evaluation requirement, multiple advertising plans can be delivered to the advertising delivery system. Otherwise, if any advertising plan does not meet the first value evaluation requirement, the first prompt information can be output.

[0196] Optionally, the output of the first prompt information may be to send the first prompt information to the user terminal, so that the user terminal displays the first prompt information in the interactive interface. The first prompt information may be used to notify the user that multiple advertising plans have failed to be verified, etc., and may also include at least one simulated advertising value corresponding to each advertising plan, so that the user can analyze the reasons and readjust the material set and / or value type, etc.

[0197] In addition, after the advertisement plan is delivered to the advertisement delivery system, its delivery effect may be tested. In some embodiments, the method may further include:

[0198] Deliver multiple advertising plans to the advertising delivery system; obtain at least one advertising value generated by any advertising plan within a predetermined time in the advertising delivery system; detect whether at least one advertising value meets a second value evaluation requirement; if not, output a second prompt message.

[0199] The advertising plan delivered to the advertising delivery system can also be tested to determine whether it meets the second value evaluation requirement. The first value evaluation requirement and the second value evaluation requirement can be the same or different, the value types to be evaluated can be the same or different, or the value types are the same but the evaluation thresholds can be different.

[0200] Advertisement plans that meet the second value evaluation requirement can continue to be delivered. If any advertisement plan does not meet the second value evaluation requirement or multiple advertisement plans do not meet the second value evaluation requirement, a second prompt message can be output.

[0201] Optionally, the output of the second prompt information may be sending the second prompt information to the user terminal so that the user terminal displays the second prompt information in the interactive interface. The second prompt information may be used to notify the user that the advertising plan delivery effect is not good, etc. In addition, it may also include at least one advertising value corresponding to the advertising plan, so that the user can perform manual evaluation to decide whether to stop delivering the multiple advertising plans, etc.

[0202] In addition, as can be seen from the above description, the estimated advertisement value can be obtained by evaluating using a value estimation model. In some embodiments, the method may further include:

[0203] Obtain a second test sample; the second test sample includes a test advertising plan and a test advertising value corresponding to the test advertising plan; use the value estimation model to estimate the estimated advertising value corresponding to the test advertising plan; determine whether the value estimation model meets the model requirements based on the estimated advertising value and the test advertising value; if not, adjust the value estimation model.

[0204] If the value estimation model meets the model requirements, the value estimation model can be used to estimate the estimated advertising value generated by the advertising plan generated based on any material in the material set.

[0205] The second test sample may also come from a historical delivery record, the test advertising plan may also be a historical advertising plan, and the test advertising value may also be a historical advertising value generated by a historical advertising plan.

[0206] The model requirement may be, for example, that the difference between the estimated advertising value and the test advertising value is within an allowable range. If the value estimation model does not meet the model requirement, the value estimation model may be adjusted, for example, returning to the step of training the value estimation model using the sample material and the sample advertising value to continue executing and retraining the value estimation model or re-adjusting the training sample.

[0207] In addition, second adjustment prompt information may be output to prompt the user whether to adjust the value estimation model, etc. The second adjustment prompt information may include difference information between the estimated advertising value and the test advertising value, etc. The second adjustment prompt information may be sent to the user terminal so that the user terminal displays the second adjustment prompt information, etc. in the interactive interface.

[0208] In the embodiment of the present application, by using the second test sample to test the value estimation model, the performance and generalization ability of the model can be evaluated to ensure the accuracy of the model.

[0209] As can be seen from the above description, the embodiments of the present application realize white-boxing of the algorithm through simulation evaluation, delivery feedback, and a visual interactive interface, etc., which enhances the controllability of the results and meets the usage requirements of upstream and downstream.

[0210] For ease of understanding, Figure 3 FIG. 1 is a conceptual diagram of an advertisement processing process in an embodiment of the present application, which may include:

[0211] Input part: The user can provide materials (target population, advertising objects and / or advertising materials), evaluation goals (value type and benefit type), constraints, total budget, etc. to the server through the interactive interface of the user terminal 201.

[0212] Advertisement generation part: The server can estimate the advertisement value based on the material set and value type in the input data, and perform optimization and solution based on the estimated advertisement value to find multiple target materials, so that multiple target materials can be used to generate corresponding multiple advertising plans. The specific implementation method can be found in the corresponding embodiment above, and will not be repeated here.

[0213] Budget allocation part: The server can estimate at least one estimated revenue value generated when any candidate budget is allocated to any one of the multiple advertising plans based on at least one revenue type in the input data, and determine the target budgets corresponding to the multiple advertising plans respectively according to the constraints and the total budget in the input data. The specific implementation method can be found in the corresponding embodiments above and will not be repeated here.

[0214] In addition, the product processing system provided in the embodiment of the present application is a white-box system, which implements algorithm processing in a white-box manner, wherein the white-box method may include a visual interactive interface, model testing, simulation testing, and delivery detection, etc., and generate corresponding prompt information as output data to interact with users to enhance the controllability of advertisements, etc. In addition, after the simulation system 203 delivers multiple advertising plans, the white-box system can also output at least one simulated revenue value and at least one simulated advertising value generated by the multiple advertising plans as output data to the user terminal 201. After the advertising delivery system 200 delivers multiple advertising plans, the white-box system can also output at least one revenue value and at least one advertising value generated by the multiple advertising plans as output data to the user terminal 201, for example. Figure 3 As shown, the white-box system can output to the user terminal 201 at least one simulated revenue value such as ROI and GMV generated by the advertising plan with ID "AAAA" placed in the simulation system 203, and display it on the interactive interface of the user terminal 201. In addition, the user's interactive results for the prompt information can be used as feedback data to make corresponding adjustments to the input data, the advertisement generation algorithm, the budget allocation algorithm and / or the user's input data.

[0215] In addition, after generating multiple advertising plans, the corresponding interface can be called to deliver them to the advertising delivery system 200, and the target budget of each advertising plan can be determined in each delivery cycle, such as every day, and updated to the advertising delivery system; the historical delivery records generated by the advertising delivery system can be used as real data to participate in the construction of the simulation system, advertising value estimation or revenue value estimation, etc.

[0216] Through the technical solution of the embodiment of the present application, multiple target materials that meet the optimization target can be found based on the materials provided by the user and the value type as the optimization target, and by estimating the advertising value, and then an advertising plan can be generated based on each target material. After the multiple advertising plans generated by the multiple target materials are put into use, the advertising value requirements can be met with a high probability, and the advertising value requirements can be pre-set in combination with the actual advertising generation requirements or the delivery effect requirements, thereby ensuring the rationality of the advertising plan generation. And by estimating the revenue value of the advertising plan and the revenue value requirements set according to the revenue type provided by the user, the target budgets corresponding to the multiple advertising plans can be determined. Under the constraint of the target budget, after the multiple advertising plans are put into use, the revenue value requirements can be met with a high probability, and the revenue value requirements can be pre-set in combination with the actual needs, thereby ensuring the rationality of the budget allocation, ensuring the advertising delivery effect, and improving the flexibility and efficiency of advertising delivery.

[0217] Among them, the algorithm is white-boxed through simulation evaluation, delivery feedback, and a visual interactive interface, which facilitates user interaction and corresponding adjustments, enhances the controllability of the results and meets the usage needs of upstream and downstream.

[0218] Figure 4 A schematic diagram of a structure of an embodiment of a budget allocation device provided in an embodiment of the present application, wherein the method and device comprises:

[0219] A first determining module 401 is used to determine a plurality of advertising plans;

[0220] An estimation module 402 is used to estimate at least one estimated revenue value generated when any advertisement plan is allocated any candidate budget;

[0221] The second determination module 403 is used to determine candidate budgets that match the multiple advertising plans respectively, taking the at least one estimated revenue value corresponding to the multiple advertising plans respectively meeting the revenue value requirement as an optimization goal;

[0222] The third determination module 404 is configured to use the candidate budgets respectively matched with the multiple advertising plans as the target budgets respectively corresponding to the multiple advertising plans.

[0223] In some embodiments, the estimating module estimating at least one estimated revenue value generated when any advertising plan is allocated to any candidate budget may include: using at least one revenue estimation model to estimate at least one estimated revenue value generated when any advertising plan is allocated to any candidate budget. Each revenue estimation model may be obtained by training based on sample advertising plans, sample budgets corresponding to the sample advertising plans, and sample revenue values.

[0224] In some embodiments, using at least one revenue estimation model, estimating at least one estimated revenue value generated when any advertising plan is allocated to any candidate budget may include: determining at least one advertising feature corresponding to any advertising plan; inputting at least one advertising feature and any candidate budget into at least one revenue estimation model to obtain at least one estimated revenue value generated when the advertising plan is allocated to the candidate budget.

[0225] In some embodiments, determining at least one advertising feature corresponding to any advertising plan may include: determining at least one average revenue value generated by any advertising plan within a historical time period; using one or more of the at least one average revenue value, population characteristics corresponding to the advertising plan, object characteristics of the advertising object corresponding to the advertising plan, and material characteristics of the advertising material corresponding to the advertising plan as at least one advertising feature corresponding to the advertising plan.

[0226] In some embodiments, the revenue estimation model can be trained in the following manner: obtain historical advertising plans and the historical budgets and historical revenue values ​​corresponding to the historical advertising plans; use the historical advertising plans as sample advertising plans, the historical budgets as sample budgets, and the historical revenue values ​​as sample revenue values; determine at least one sample advertising feature of the sample advertising plan; use at least one sample advertising feature of the sample advertising plan and the sample budget as input data, and the sample revenue values ​​as training labels to train the revenue estimation model.

[0227] In some embodiments, the device can also be used to: obtain a first test sample; the first test sample includes a test advertising plan and a test revenue value corresponding to the test advertising plan; use a revenue estimation model to estimate the estimated revenue value corresponding to the test advertising plan; determine whether the revenue estimation model meets the model requirements based on the estimated revenue value and the test revenue value; if not, adjust the revenue estimation model.

[0228] In addition, the device can also output first adjustment prompt information to prompt the user whether to adjust the profit estimation model, etc.

[0229] In some embodiments, the device may also determine budget candidate sets corresponding to a plurality of advertising plans.

[0230] The estimating module estimating at least one estimated revenue value generated when any advertising plan is allocated to any candidate budget may include: estimating at least one estimated revenue value generated when any advertising plan is allocated to any candidate budget in its corresponding budget candidate set.

[0231] In some embodiments, determining budget candidate sets corresponding to multiple advertising plans may include: for any advertising plan, based on the actual budget consumed by the advertising plan in the previous delivery cycle, generating multiple candidate budgets within the budget range corresponding to the actual budget to form a budget candidate set.

[0232] The third determination module uses the candidate budgets matched to the multiple advertising plans as the target budgets corresponding to the multiple advertising plans respectively, which may include: using the candidate budgets matched to the multiple advertising plans as the target budgets corresponding to the multiple advertising plans in the current delivery cycle respectively.

[0233] The device can also be used to update the target budgets corresponding to multiple advertising plans in the current delivery cycle to the advertising delivery system, so that the advertising delivery system can display advertisements for multiple advertising plans according to their corresponding target budgets in the current delivery cycle.

[0234] In some embodiments, the second determination module takes at least one estimated revenue value corresponding to each of the multiple advertising plans satisfying the revenue value requirement as an optimization goal, and determining the candidate budgets that match the multiple advertising plans may include: taking the total value corresponding to the first estimated revenue values ​​generated by the multiple advertising plans as the maximum as the optimization goal, at least one second estimated revenue value generated by the multiple advertising plans as being within the revenue limit range, and the total budget of the candidate budgets corresponding to the multiple advertising plans as being within the budget limit range as constraints, to determine the candidate budgets that match the multiple advertising plans.

[0235] In some embodiments, the device can also be used to: obtain at least one revenue value generated by any advertising plan within a predetermined time in the advertising delivery system; detect whether at least one revenue value meets a first revenue evaluation requirement; if not, use the actual budget of the advertising plan in the previous delivery cycle as the target budget for the current delivery cycle, and update it to the advertising delivery system.

[0236] In some embodiments, the device may also be used to: provide an interactive interface to the user terminal; and determine multiple advertising plans provided by the user based on interactive information sent by the user terminal.

[0237] In some embodiments, the estimating module estimating at least one estimated revenue value generated when any advertising plan is allocated to any candidate budget may include: estimating at least one estimated revenue value generated when any advertising plan is allocated to any candidate budget based on at least one revenue type.

[0238] In some embodiments, the device may also send advertisement selection prompt information and revenue type prompt information to the user terminal.

[0239] In some embodiments, the device can also deliver multiple advertising plans in the simulation system according to the target budget, and obtain at least one simulated revenue value generated by the multiple advertising plans; determine whether the at least one simulated revenue value meets the second revenue evaluation requirement; if so, deliver the multiple advertising plans to the advertising delivery system; if not, output a warning prompt message.

[0240] In some embodiments, the device can also: determine a material set; estimate at least one estimated advertising value generated by an advertising plan generated based on any material in the material set; search for multiple target materials so that at least one estimated advertising value respectively generated by multiple advertising plans generated based on multiple target materials meets the advertising value requirement; and generate multiple advertising plans using multiple target materials.

[0241] In some embodiments, estimating at least one first estimated advertising value generated by an advertising plan generated based on any one of the material set may include: using at least one value estimation model to estimate at least one estimated advertising value generated by an advertising plan generated based on any one of the material set. The value estimation model may be obtained by training based on sample materials and sample advertising values ​​corresponding to the sample materials.

[0242] In some embodiments, the value estimation model can be pre-trained in the following manner:

[0243] Determine the historical advertising plan and the historical advertising value generated by the historical advertising plan; use the materials involved in the historical advertising plan as sample materials, and the historical advertising value as sample advertising value; use the sample materials and sample advertising value to train the value estimation model.

[0244] In some embodiments, the device may also be used to: perform weighted processing on at least one estimated advertising value corresponding to any material to generate a material value corresponding to the material.

[0245] After obtaining the material value, in one implementation method, the above-mentioned searching for multiple target materials so that at least one estimated advertising value respectively generated by multiple advertising plans generated based on the multiple target materials meets the advertising value requirement can be: searching for multiple target materials so that the material values ​​corresponding to the multiple target materials respectively meet the advertising value requirement.

[0246] As another implementation, searching for multiple target materials so that at least one estimated advertising value generated by multiple advertising plans generated based on the multiple target materials meets the advertising value requirement can be: taking the material value difference between materials in the same advertising plan to meet the difference requirement as the optimization goal, searching for at least two target materials that match the multiple advertising plans. This implementation can ensure the homogeneity of at least two target materials in the multiple advertising plans.

[0247] In addition, as another implementation method, the materials in the material set can be grouped in different division methods, and a corresponding advertising plan can be generated by using the materials in the same group as the target materials. The advertising plan corresponding to each group can be considered as an advertising plan that meets the advertising value requirements.

[0248] In addition, to ensure delivery efficiency and delivery effects, the advertising value requirements may include optimization objectives and constraints. The optimization objective may, for example, be that the differences between one or more estimated advertising values ​​corresponding to multiple advertising plans generated by multiple target materials meet the difference requirements. The constraints may, for example, include that the number of advertising plans is the target number, the sum of the estimated values ​​of an advertisement corresponding to multiple advertising plans is within a predetermined value range, etc.

[0249] In some embodiments, the above-mentioned searching for multiple target materials so that at least one estimated advertising value respectively generated by multiple advertising plans generated based on the multiple target materials meets the advertising value requirement may include:

[0250] With the first estimated advertising values ​​respectively generated by multiple advertising plans satisfying the advertising value requirements as the optimization goal, and the second estimated advertising values ​​respectively generated by multiple advertising plans or the sum of the second estimated advertising values ​​being within a predetermined value range as the constraint conditions, the target materials matching the multiple advertising plans are obtained.

[0251] In some embodiments, the device may also: provide an interactive interface to the user terminal so that the user can provide a material set in the interactive interface; and obtain the material set sent by the user terminal.

[0252] The above-mentioned at least one estimated advertising value generated by an advertising plan generated based on any material in the material set may include: at least one estimated advertising value corresponding to at least one advertising value type generated by an advertising plan generated based on any material in the material set.

[0253] In some possible implementations, the device may send material prompt information to the user terminal so that the user terminal displays the material prompt information in the interactive interface. In addition, the device may send value type prompt information to the user so that the user terminal displays the value type prompt information in the interactive interface.

[0254] In some embodiments, the device can also: place multiple advertising plans in the simulation system and obtain at least one simulated advertising value generated by each of the multiple advertising plans; determine whether the at least one simulated advertising value generated by each of the multiple advertising plans meets the first value evaluation requirement; if so, place the multiple advertising plans in the advertising delivery system; if not, output a first prompt message.

[0255] In addition, after the advertising plan is delivered to the advertising delivery system, its delivery effect can also be detected. In some embodiments, the device can also: deliver multiple advertising plans to the advertising delivery system; obtain at least one advertising value generated by any advertising plan within a predetermined time in the advertising delivery system; detect whether at least one advertising value meets the second value evaluation requirement; if not, output a second prompt message.

[0256] The advertising plan delivered to the advertising delivery system can also be tested to determine whether it meets the second value evaluation requirement. The first value evaluation requirement and the second value evaluation requirement can be the same or different, the value types to be evaluated can be the same or different, or the value types are the same but the evaluation thresholds can be different.

[0257] Advertisement plans that meet the second value evaluation requirement can continue to be delivered. If any advertisement plan does not meet the second value evaluation requirement or multiple advertisement plans do not meet the second value evaluation requirement, a second prompt message can be output.

[0258] In addition, as can be seen from the above description, the estimated advertising value can be obtained by evaluating using a value estimation model. In some embodiments, the device can also: obtain a second test sample; the second test sample includes a test advertising plan and a test advertising value corresponding to the test advertising plan; use the value estimation model to estimate the estimated advertising value corresponding to the test advertising plan; determine whether the value estimation model meets the model requirements based on the estimated advertising value and the test advertising value; if not, adjust the value estimation model.

[0259] If the value estimation model meets the model requirements, the value estimation model can be used to estimate the estimated advertising value generated by the advertising plan generated based on any material in the material set.

[0260] The second test sample may also come from a historical delivery record, the test advertising plan may also be a historical advertising plan, and the test advertising value may also be a historical advertising value generated by a historical advertising plan.

[0261] In addition, the device can also output a second adjustment prompt information to prompt the user whether to adjust the value estimation model, etc. Figure 4 The budget allocation device can execute Figure 1 The implementation principle and technical effect of the budget allocation method described in the embodiment are not described in detail. The specific way in which each module and unit performs operations in the budget allocation device in the above embodiment has been described in detail in the embodiment of the method, and will not be described in detail here.

[0262] The present application also provides a computing device, such as Figure 5 As shown, the device may include a storage component 501 and a processing component 502;

[0263] The storage component 501 stores one or more computer instructions, wherein the one or more computer instructions are called and executed by the processing component to implement the following Figure 1 The budget allocation method described in the illustrated embodiment.

[0264] Of course, the computing device may also include other components, such as input / output interfaces, display components, communication components, etc.

[0265] The input / output interface provides an interface between the processing component and the peripheral interface module, which may be an output device, an input device, etc. The communication component is configured to facilitate wired or wireless communication between the computing device and other devices.

[0266] The processing component 502 may include one or more processors to execute computer instructions to complete all or part of the steps in the above method. Of course, the processing component may also be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors or other electronic components to perform the above method.

[0267] The storage component 501 is configured to store various types of data to support operations at the terminal. The storage component can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0268] The display component may be an electroluminescent (EL) element, a liquid crystal display or a micro display having a similar structure, or a retinal direct display or a similar laser scanning display.

[0269] It should be noted that the above computing device can be a physical device or an elastic computing host provided by a cloud computing platform, etc. It can be implemented as a distributed cluster consisting of multiple servers or terminal devices, or as a single server or a single terminal device.

[0270] It should be noted that the above computing device implements Figure 1In the case of the budget allocation method described in the illustrated embodiment, it can be a physical device or an elastic computing host provided by a cloud computing platform, etc. It can be implemented as a distributed cluster consisting of multiple servers or terminal devices, or as a single server or a single terminal device.

[0271] The present application also provides a computer-readable storage medium storing a computer program, wherein the computer program can achieve the above-mentioned Figure 1 The budget allocation method described in the illustrated embodiment. The computer-readable medium may be included in the electronic device described in the above embodiment; or it may exist independently without being assembled into the electronic device.

[0272] The present application also provides a computer program product, which includes a computer program carried on a computer-readable storage medium. When the computer program is executed by a computer, the computer program can achieve the above-mentioned Figure 1 The budget allocation method described in the illustrated embodiment. In such an embodiment, the computer program can be downloaded and installed from a network, and / or installed from a removable medium. When the computer program is executed by a processor, various functions defined in the system of the present application are executed.

[0273] In the above corresponding embodiments, the computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a flash memory, an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the present application, a computer-readable storage medium may be any tangible medium containing or storing a program that may be used by or in combination with an instruction execution system, device, or device.

[0274] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0275] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0276] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0277] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A budget allocation method, characterized in that: include: Identify multiple advertising plans; estimating at least one estimated revenue value generated by allocating any candidate budget to any advertising plan; Taking at least one estimated revenue value corresponding to each of the multiple advertising plans satisfying a revenue value requirement as an optimization goal, determining candidate budgets that match each of the multiple advertising plans; The candidate budgets respectively matched to the multiple advertising plans are used as the target budgets respectively corresponding to the multiple advertising plans.

2. The method according to claim 1, characterized in that The at least one estimated revenue value generated when any advertising plan is allocated any candidate budget includes: Using at least one revenue estimation model, estimating at least one estimated revenue value generated when any advertising plan is allocated with any candidate budget; The revenue estimation model is obtained by training based on a sample advertising plan, a sample budget corresponding to the sample advertising plan, and a sample revenue value.

3. The method according to claim 2, characterized in that The method of using at least one revenue estimation model to estimate at least one estimated revenue value generated when any advertising plan is allocated any candidate budget includes: Determining at least one advertisement characteristic corresponding to any advertisement plan; The at least one advertisement feature and any candidate budget are input into at least one revenue estimation model to obtain at least one estimated revenue value generated when the candidate budget is allocated to the advertisement plan.

4. The method according to claim 3, characterized in that The determining of at least one advertisement feature corresponding to any advertisement plan comprises: Determine at least one average revenue value generated by any advertising program over a historical time period; One or more of the at least one average revenue value, the population characteristics corresponding to the advertising plan, the object characteristics of the advertising object corresponding to the advertising plan, and the material characteristics of the advertising material corresponding to the advertising plan are used as at least one advertising feature corresponding to the advertising plan.

5. The method according to claim 3, characterized in that: The revenue estimation model is trained in the following way: Obtaining historical advertising plans and historical budgets and historical revenue values ​​corresponding to the historical advertising plans; using the historical advertising plan as a sample advertising plan, the historical budget as a sample budget, and the historical revenue value as a sample revenue value; determining at least one sample advertising characteristic of the sample advertising campaign; At least one sample advertisement feature of the sample advertisement plan and the sample budget are used as input data, and the sample revenue value is used as a training label to train a revenue estimation model.

6. The method according to claim 1, characterized in that Also includes: Determine budget candidate sets corresponding to the multiple advertising plans respectively; The at least one estimated revenue value generated when any advertising plan is allocated any candidate budget includes: At least one estimated revenue value generated when any advertisement plan allocates any candidate budget in the corresponding preset candidate set is estimated.

7. The method according to claim 6, characterized in that Determining the budget candidate sets corresponding to the multiple advertising plans respectively includes: For any advertising plan, according to the actual budget consumed by the advertising plan in the last delivery cycle, within the budget range corresponding to the actual budget, multiple candidate budgets are generated to form the budget candidate set; The candidate budgets respectively matching the multiple advertising plans as the target budgets respectively corresponding to the multiple advertising plans include: Using the candidate budgets respectively matched by the multiple advertising plans as the target budgets respectively corresponding to the multiple advertising plans in the current delivery cycle; The method further comprises: The target budgets respectively corresponding to the multiple advertising plans in the current delivery cycle are updated to the advertising delivery system, so that the advertising delivery system can perform promotion operations for the multiple advertising plans in the current delivery cycle according to their respective corresponding target budgets.

8. The method according to claim 1, characterized in that The step of determining the candidate budgets that the plurality of advertising plans are matched with each other by taking at least one estimated revenue value corresponding to each of the plurality of advertising plans satisfying the revenue value requirement as the optimization goal comprises: The candidate budgets that match the multiple advertising plans are determined by taking the maximization of the total value corresponding to the first estimated revenue values ​​respectively generated by the multiple advertising plans as the optimization goal, at least one second estimated revenue value respectively generated by the multiple advertising plans being within the revenue limit range, and the total budget of the candidate budgets respectively corresponding to the multiple advertising plans being within the budget limit range as the constraint conditions.

9. The method according to claim 1, characterized in that: Also includes: updating the target budgets respectively corresponding to the multiple advertising plans to the advertising delivery system, so that the advertising delivery system performs promotion operations for the multiple advertising plans according to the target budgets respectively corresponding to the multiple advertising plans; Obtain at least one revenue value generated by any advertisement plan in the advertisement delivery system within a predetermined time; detecting whether the at least one benefit value satisfies a first benefit assessment requirement; If not, the actual budget of the previous delivery cycle of the advertising plan is used as the target budget of the current delivery cycle and updated to the advertising delivery system.

10. The method according to claim 1, characterized in that Also includes: Delivering the multiple advertising plans according to the target budget in the simulation system, and obtaining at least one simulation revenue value generated by the multiple advertising plans; determining whether the at least one simulation benefit value satisfies a second benefit assessment requirement; If so, delivering the plurality of advertisement plans to an advertisement delivery system; If not, output a warning message.

11. A computing device, characterized in that: It comprises a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement the budget allocation method as claimed in claim 1.

12. A computer storage medium, characterized in that: A computer program is stored, and when the computer program is executed by a computer, the budget allocation method according to claim 1 is implemented.

13. A computer program product, characterized in that The method comprises a computer program / instruction, and when the computer program / instruction is executed by a computer, the budget allocation method as claimed in claim 1 is implemented.

Images