An advertising delivery method, apparatus, storage medium, electronic device, and product.

Abstract

This specification provides an advertising delivery method, apparatus, storage medium, electronic device, and product. The method includes: obtaining ad placement information and corresponding candidate creatives for the current ad placement page based on a user-triggered action; wherein the candidate creatives include budget-controlled creatives subject to budget constraints; determining the current budget consumption data for the ad placement information, and for each budget-controlled creative, obtaining a matching score between the creative and the user; determining a bid adjustment coefficient based on the deviation between the current budget consumption data and the target budget consumption status; determining a dynamic recommendation score for the budget-controlled creative based on the matching score and the adjusted bid based on the bid adjustment coefficient; and determining target creatives based on the dynamic recommendation scores of each budget-controlled creative and delivering them to the user.

Description

Technical Field

[0001] This specification relates to one or more embodiments in the field of computer technology, and more particularly to an advertising delivery method, apparatus, storage medium, electronic device, and product. Background Technology

[0002] In internet advertising scenarios, it's typically necessary to select and deliver ads from multiple candidate creatives when a user visits a page, enters an application, or triggers an interactive action. As advertising platforms handle an increasing variety of business types, the ad delivery process often requires comprehensive consideration of multiple factors. It needs to complete ad selection within a short processing time while balancing performance and budget execution requirements.

[0003] However, existing advertising delivery solutions typically distribute ads according to preset delivery rules or coarse-grained budget adjustments. While these solutions can achieve basic ad delivery and budget management functions, they suffer from poor stability of delivery results and difficulty in simultaneously achieving both performance and budget targets. Summary of the Invention

[0004] In view of the above, one or more embodiments of this specification provide the following technical solutions: According to a first aspect of one or more embodiments of this specification, an advertising delivery method is provided, comprising: Based on the user's execution of the campaign trigger operation, obtain the booth information of the current campaign page and the candidate creatives corresponding to the booth information; wherein, the candidate creatives include budget-controlled creatives subject to budget constraints; Determine the current budget consumption data of the booth information, and for each budget control creative, obtain the matching score between the budget control creative and the user; The bid adjustment factor is determined based on the deviation between the current budget consumption data and the target budget consumption status; Based on the matching score and the creative bid adjusted based on the bid adjustment coefficient, the dynamic recommendation score corresponding to the budget-controlled creative is determined; Target creatives are determined based on the dynamic recommendation scores corresponding to each budget-controlled creative, and then delivered to the users.

[0005] According to a second aspect of one or more embodiments of this specification, an advertising delivery device is provided, comprising: The acquisition unit acquires the booth information of the current ad placement page and the candidate creatives corresponding to the booth information based on the ad placement trigger operation performed by the user; wherein, the candidate creatives include budget-controlled creatives subject to budget constraints. The unit determines the current budget consumption data of the booth information and, for each budget control creative, obtains the matching score between the budget control creative and the user. The calculation unit determines the bid adjustment coefficient based on the deviation between the current budget consumption data and the target budget consumption status; The adjustment unit determines the dynamic recommendation score corresponding to the budget-controlled creative based on the matching score and the creative bid adjusted based on the bid adjustment coefficient. The delivery unit determines the target creative based on the dynamic recommendation score corresponding to each budget-controlled creative and delivers it to the user.

[0006] According to a third aspect of one or more embodiments of this specification, an electronic device is provided, comprising: a processor; a memory for storing processor-executable instructions; wherein the processor implements the steps of the method described above by executing the executable instructions.

[0007] According to a fourth aspect of one or more embodiments of this specification, a computer-readable storage medium is provided that stores computer instructions thereon, which, when executed by a processor, implement the steps of the method described above.

[0008] According to a fifth aspect of one or more embodiments of this specification, a computer program product is provided, comprising a computer program / instructions that, when executed by a processor, implement the steps of the method described above.

[0009] As described in the above embodiments, when a user triggers an ad delivery request, this specification obtains the ad placement information and corresponding candidate creatives on the current delivery page. For budget-controlled creatives, it determines the bid adjustment coefficient by combining the deviation between the current budget consumption data and the target budget consumption status, and determines the dynamic recommendation score based on the matching score and the adjusted creative bid. Through this method, the recommendation results of budget-controlled creatives can be adapted to the current budget execution status, improving the timeliness and precision of budget adjustment, reducing the inconsistency between the budget consumption process and the actual delivery rhythm, further enhancing the rationality of target creative determination and the stability of delivery results, and improving the overall performance of the ad delivery system in complex business scenarios. Attached Figure Description

[0010] Figure 1 This is a schematic diagram of the architecture of an advertising delivery service system provided in an exemplary embodiment; Figure 2 This is a flowchart illustrating an exemplary embodiment of an advertising delivery method; Figure 3 This is an exemplary embodiment of an overall flowchart for ad delivery; Figure 4 This is a logical architecture diagram of an exemplary embodiment for ad delivery; Figure 5 This is a schematic diagram of the structure of a device provided in an exemplary embodiment; Figure 6 This is a block diagram of an advertising delivery device provided in an exemplary embodiment. Detailed Implementation

[0011] To enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this specification, and not all embodiments. Based on the embodiments in this specification, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this specification.

[0012] 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, data stored, data displayed, etc.) involved in this manual are all information and data authorized by the user or fully authorized by all parties. The collection, use and processing of related data shall comply with the relevant laws, regulations and standards of the relevant regions, and corresponding operation portals shall be provided for users to choose to authorize or refuse.

[0013] In internet advertising scenarios, platforms typically need to select suitable target creatives from multiple candidate creatives within a short period of time when users visit a page, enter an application, or trigger an interactive action. As advertising business expands and the types of ad placements increase, the ad placement process is no longer simply a matter of creative ranking; it is also affected by factors such as budget execution pace, differences in creative bids, user response tendencies, ad placement traffic fluctuations, and system processing performance. Especially under conditions of continuous high-concurrency requests, if static recommendations or coarse-grained budget adjustments are still used, problems can easily arise such as the budget being consumed too quickly in certain periods, preventing continued ad placement during subsequent high-traffic periods, or excessively tight budget control preventing high-quality creatives that should have received exposure from being displayed.

[0014] In addition, existing advertising solutions typically have the following shortcomings in terms of engineering implementation: Budget adjustments often rely on fixed rules or low-frequency adjustment mechanisms, which are not timely enough to respond to the current budget execution status and are difficult to adapt to dynamic scenarios such as sudden increases in traffic, decreases in traffic, or changes in booth competition. Some solutions only control at the planning or unit level, resulting in coarse-grained control that cannot fine-tune the differences in creativity. When the volume of ad requests is large, multiple external data source queries, sorting calculations and filtering logic can easily introduce additional latency. If the main process takes too long, the delivery results may fall back to the default logic, affecting the stability of the delivery. Existing solutions often struggle to simultaneously balance budget smoothing and ad performance. When budget control is tight, ad performance tends to decline, while prioritizing performance can lead to budget execution deviating from the target schedule. For these reasons, an ad delivery method is needed that can balance budget execution and recommendation effectiveness in various ad delivery scenarios.

[0015] Based on this, this specification provides an advertising delivery method. After a user triggers a delivery request, the method obtains the ad placement information of the current delivery page and the corresponding candidate creatives, distinguishing between budget-controlled and non-budget-controlled creatives. Then, it determines the current budget consumption data for the ad placement information and obtains the matching score between the budget-controlled creative and the user. Further, based on the deviation between the current budget consumption data and the target budget consumption status, combined with the historical cumulative amount and trend of the deviation, it determines a bid adjustment coefficient. Then, based on the matching score and the creative bid adjusted according to the bid adjustment coefficient, it determines the dynamic recommendation score corresponding to the budget-controlled creative. Finally, based on the dynamic recommendation score corresponding to each budget-controlled creative and the preset recommendation score corresponding to each non-budget-controlled creative, it determines the target creative and delivers it to the user. This processing method allows the ranking of budget-controlled creatives to dynamically change with the budget execution status, maintaining the real-time nature of advertising delivery while improving the coordination between budget execution and delivery effectiveness.

[0016] To facilitate understanding of the technical solutions in this specification, some of the concepts involved in this specification are explained below.

[0017] Booth Information: Information related to the location of the ad creative displayed on the current ad placement page. Booth information may include at least one of the following: booth identifier, booth type, page location identifier, business page identifier, and budget configuration identifier corresponding to the booth. Different booth information can correspond to different budget control strategies, candidate creative sets, and display rules.

[0018] Candidate Creatives: The set of creatives available for selection and placement by the system in the current ad request. Candidate creatives can include various forms such as image creatives, copy creatives, card creatives, image and text combination creatives, and landing page entry creatives. Based on whether they are subject to budget control, candidate creatives can be divided into budget-controlled creatives and non-budget-controlled creatives. Budget-controlled creatives are those subject to budget constraints during ad placement (i.e., requiring budget resources). These creatives are typically used to offer discounts, coupons, benefits, or other content with budget consumption attributes and are usually categorized under budget units. Non-budget-controlled creatives are those not subject to budget constraints during ad placement (i.e., not consuming budget resources). These creatives primarily aim to gain ad placement exposure and are usually categorized under competitive units.

[0019] Current budget consumption data: The actual consumption of the budget associated with the current booth information at the current moment or within the current statistical window. Current budget consumption data can be the cumulative consumption value, the consumption value per unit time, or the segmented consumption value divided by time slices. This data can be provided by a real-time budget service, a caching system, or a log aggregation system.

[0020] Target Budget Consumption Status: This describes the expected consumption status of the budget within the current campaign period. The target budget consumption status can include the budget ceiling and the target budget consumption value corresponding to at least one time point, representing the expected execution progress of the budget at different time points. The target budget consumption status can be generated based on the booth budget configuration, campaign period configuration, traffic forecast results, or historical traffic trends.

[0021] Match Score: A quantitative result of the fit between budget-controlled creatives and users. This score can be used to characterize the likelihood of a user clicking, converting, staying, or engaging in other targeted behaviors with the budget-controlled creative. The match score can be predicted by a model or calculated based on user historical behavior data, creative attribute data, and contextual features.

[0022] Bid Adjustment Factor: A factor used to adjust creative bids based on budget execution status. This factor typically reflects the degree of deviation between the current budget expenditure and the target budget expenditure. When actual budget expenditure is lower than the target progress, the bid adjustment factor can be increased to improve creative competitiveness; when actual budget expenditure is higher than the target progress, the bid adjustment factor can be decreased to curb further accelerated budget expenditure.

[0023] Dynamic Recommendation Score: This is the real-time recommendation score used by the budget-controlled creative when determining its suitability in the current ad request scenario. This score is not a fixed value but is dynamically determined based on factors such as the match score and adjusted creative bids, thus reflecting both user matching and budget control needs.

[0024] Preset Recommendation Score: This is a pre-configured recommendation score for non-budget-controlled creatives or a score obtained directly from existing recommendation logic. Since non-budget-controlled creatives do not participate in the bid adjustment process driven by the current budget status, this preset recommendation score can be directly used in determining the target creative.

[0025] The technical solutions described in the embodiments of this specification will be explained in detail below with reference to the accompanying drawings: Figure 1 This is a schematic diagram of the architecture of an advertising delivery service system provided in an exemplary embodiment. For example... Figure 1 As shown, the system may include a server 11, a network 12, and several electronic devices, such as a personal computer (PC) 13, a mobile phone 14, etc.

[0026] Server 11 can be a physical server containing an independent host, or it can be a virtual server hosted in a host cluster. During operation, server 11 can run server-side programs for a certain application to implement the relevant functions of that application. For example, when server 11 runs an advertising delivery service program, it can function as a corresponding advertising delivery service platform.

[0027] PC13 and mobile phone14 are just some of the types of electronic devices that users can use. In reality, users can obviously also use electronic devices such as tablets, laptops, PDAs (Personal Digital Assistants), wearable devices (such as smart glasses, smartwatches, etc.), etc., and one or more embodiments in this specification do not limit this. During operation, the electronic device can run a client-side program of an application to achieve the relevant functions of that application. For example, when the electronic device runs an advertising service program, it can act as a client for that advertising service. The aforementioned advertising service client application can be launched and run on the electronic device. This client-side program can be a native application installed on the electronic device, or it can be a mini-program, quick app, or other similar form. Of course, when using web technologies such as HTML5 or similar, the relevant functions can be achieved through a page displayed by a browser. This browser can be a standalone browser application or a browser module embedded in some applications.

[0028] As for the network 12 that enables interaction between electronic devices such as PC13 and mobile phone 14 and server 11, communication can be achieved using either wired or wireless networks, depending on the communication methods supported by the respective electronic devices. This specification does not impose any restrictions on this. For example, PC13 can support both wired and wireless communication, so it can use either wired or wireless networks as needed. Mobile phone 14 typically only supports wireless communication, so it can use a wireless network for communication.

[0029] In this specification, the entity executing the advertising delivery method can be an advertising delivery platform, specifically a server of the advertising delivery platform or other electronic devices with data processing capabilities. It should be understood that a server can be a single physical server, a virtual machine, a container cluster, a cloud service node, or a distributed system composed of multiple computing nodes. For ease of description, the following will use a server as the executing entity to describe the advertising delivery method provided in this specification.

[0030] Figure 2 This is a flowchart illustrating an exemplary embodiment of an advertising delivery method, including the following steps: S200: Based on the user's execution of the campaign trigger operation, obtain the booth information of the current campaign page and the candidate creative corresponding to the booth information; wherein, the candidate creative includes budget-controlled creative subject to budget constraints.

[0031] In practical applications, actions that trigger ad delivery can include a user opening a page in the client or webpage, pulling down to refresh the page, clicking an entry point, switching business tabs, entering a details page, or other actions that trigger an ad delivery request. After receiving an ad delivery request, the server can parse the page identifier, user identifier, device identifier, business context information, and ad placement request parameters in the request, and determine the ad placement information that needs to be processed.

[0032] Once the booth information is determined, the server can retrieve the candidate creatives corresponding to that booth information from the creative retrieval service, the delivery configuration service, or the local cache.

[0033] In addition, candidate creative ideas corresponding to booth information can be obtained through various methods. For example: Based on the booth information, read the set of creative ideas bound to the booth from the pre-configured creative pool; Based on booth information, user characteristics, and business context, creative ideas are recalled from multiple sources and merged to form a set of candidate creative ideas; First, obtain the creatives that are currently in the deployment state under this booth. Then, perform preliminary filtering based on the creative validity period, review status, deployment status, and business adaptation rules to obtain the candidate creatives for the current request.

[0034] The server can pre-classify each candidate creative based on the creative's delivery unit (such as budget unit, competition unit, etc.), budget configuration identifier, or budget management attribute, and divide them into budget-controlled creatives and non-budget-controlled creatives.

[0035] In one optional embodiment, after performing step S200, the server may further obtain the user's activity status data. This activity status data may be determined based on at least one of the following: access frequency, page view frequency, application launch frequency, historical interaction records, or business activity tags within a preset time period.

[0036] In practical applications, for highly active users, the additional revenue generated by continuing to execute the complete advertising delivery process for them is relatively limited, since these users have a high frequency of visits and high business participation. For inactive users, however, adopting a dynamic delivery strategy is more conducive to improving advertising reach and budget utilization efficiency.

[0037] Therefore, if a user is determined to be a highly active user based on activity status data, the server can either target that user with ads or deliver ad creatives to that user that are not subject to budget control. If the user is determined to be a low-activity user based on the activity status data, the server continues to execute the target creative determination process in steps S202~S208.

[0038] S202: Determine the current budget consumption data for the booth information, and for each budget-controlled creative, obtain the matching score between the budget-controlled creative and the user.

[0039] Based on the booth information, the server can retrieve the current budget consumption data associated with that booth information from the real-time budget service, budget cache, or budget data aggregation module.

[0040] In practical applications, the current budget consumption data can be the cumulative consumption value for the current calendar day, the cumulative consumption value for the current campaign period, or the consumption value up to the current time point. If the booth budget is bound to an experimental version, traffic bucket, or business line, the server can further locate the corresponding budget instance and determine the current budget consumption data of that budget instance by combining the experimental identifier, traffic bucket identifier, and business identifier.

[0041] At the same time, for each budget-controlled ad, the server can also obtain a matching score between the ad and the user. The matching score can be output by the recommendation model or calculated using rules. For example, the server can generate a matching score based on at least one of the following: user historical exposure data, historical click data, historical conversion data, historical dwell time, ad category preferences, ad placement context features, and time context features.

[0042] In this specification, the matching score between budget control ideas and users can be pre-stored in a cache. This cache can include a first-level cache and a second-level cache. The first-level cache can include local memory to store the matching scores corresponding to cache index identifiers accessed more frequently than a preset frequency; the second-level cache can include a distributed cache to store the matching scores corresponding to a wider range of cache index identifiers.

[0043] During the process of obtaining the matching score, the server can determine the cache index identifier based on the creative identifier of the budget control creative and the user's user identifier, and retrieve the corresponding matching score in the preset first-level cache based on the cache index identifier.

[0044] Since the first-level cache is typically implemented in local memory, it offers faster access speeds. Therefore, the server can prioritize retrieving the corresponding matching score from the first-level cache, thereby improving the efficiency of matching score retrieval. If the matching score is not found in the first-level cache, the server can further retrieve the matching score corresponding to that cache index from the second-level cache.

[0045] In addition, upon reaching the preset detection time, the server can also obtain the cache timestamps corresponding to each cache index identifier in the first-level cache and second-level cache, and calculate the time interval between the preset detection time and the cache timestamps corresponding to each cache index identifier. If the time interval corresponding to any cache index identifier exceeds the preset time threshold, the server can update the matching score corresponding to that cache index identifier, thereby ensuring that the matching score in the cache has good timeliness and improving the accuracy of subsequent matching score acquisition results.

[0046] The matching score can be updated by re-fetching user feature data (such as user historical clicks, user historical browsing history, user interest preference tags, etc.) and creative feature data, and then recalculating the matching score based on the updated user feature data and creative feature data. The preset detection time can be set based on a cache refresh strategy, such as a detection time triggered at a fixed time interval, or a detection time corresponding to off-peak business periods.

[0047] In addition, if neither the first-level cache nor the second-level cache hits the corresponding matching score, the server can call the recommendation model service in real time to obtain the matching score between the budget-controlled creative and the user, and then write the matching score into the first-level cache and / or the second-level cache for reuse in subsequent requests.

[0048] Additionally, if user feature data and / or creative feature data for calculating the matching score are unavailable, the server may be unable to calculate the matching score between the budget-controlled creative and the user in real time (e.g., an intermediate matching score). In this case, the server can use a preset default score as the matching score corresponding to the budget-controlled creative to ensure that the subsequent dynamic recommendation score determination process can continue.

[0049] In addition, the server can also execute creative filtering logic. This filtering logic can be executed after step S202 and before step S204, or it can be executed in combination with other steps.

[0050] Specifically, the server can determine the target fatigue score for each candidate creative and filter out candidate creatives whose target fatigue score is greater than the preset score, in order to reduce user fatigue from repetitive creatives.

[0051] In this specification, the server can determine fatigue-related scores from business, booth, and creative dimensions. Specifically, the server can determine the business dimension fatigue score based on the historical recommendation data of the entire business chain to users; determine the booth dimension fatigue score based on the historical recommendation data of the ad placement page to users; determine the creative dimension fatigue score for each candidate creative based on the historical recommendation data of each candidate creative to users; and then, for each candidate creative, determine the target fatigue score corresponding to that candidate creative based on the business dimension fatigue score, the booth dimension fatigue score, and the creative dimension fatigue score corresponding to that candidate creative.

[0052] The scores across the three dimensions can be combined using methods such as weighted summation, linear combination, tiered threshold judgment, or maximum value selection. For example, the target fatigue score can be represented as the sum of the fatigue score in the business dimension multiplied by the first weight, the fatigue score in the booth dimension multiplied by the second weight, and the fatigue score in the creative dimension multiplied by the third weight. If the target fatigue score is greater than a preset score, the server will filter out the corresponding candidate creative from the candidate set.

[0053] It's worth noting that the matching score can also be determined using federated learning technology. While protecting user privacy and data security, the server can collaborate with data nodes from multiple business teams and / or multiple client devices to train a recommendation model based on user behavior characteristics, traffic characteristics, and ad placement strategy characteristics related to ad delivery. This results in a model used to determine the matching score. This approach improves the accuracy of matching score calculations and the model's generalization ability by leveraging multi-source data without directly aggregating raw data.

[0054] S204: Determine the bid adjustment coefficient based on the deviation between the current budget consumption data and the target budget consumption status; S206: Based on the matching score and the creative bid adjusted based on the bid adjustment coefficient, determine the dynamic recommendation score corresponding to the budget-controlled creative; The server can first determine the target budget consumption status corresponding to the current booth. The target budget consumption status can include the budget ceiling and the target budget consumption value at least at one time point. The budget ceiling can be determined based on the booth's basic budget configuration, or it can be adjusted based on the basic budget configuration and historical traffic trends.

[0055] For example, the server can adjust the base budget cap for booth information based on historical traffic trends to obtain the adjusted budget cap. Historical traffic trends can be generated based on at least one of the following metrics for the same time period in historical dates: request volume, impressions, click-through rate, conversion rate, competition intensity, or traffic quality. When it is detected that the current time period typically has high traffic value, the server can appropriately increase the adjusted budget cap; when it is detected that the current time period has low traffic value or that recent budgets have been executed too quickly, the server can appropriately decrease the adjusted budget cap.

[0056] After receiving the adjusted budget cap, the server can determine the target budget consumption value corresponding to the current time point.

[0057] For example, if the campaign period is one day, the server can divide the day into multiple time nodes according to preset time slices and configure the target budget consumption value for each time node; it can also generate a non-uniform budget rhythm based on historical traffic distribution, so that high-value periods correspond to higher target budget consumption values ​​and low-value periods correspond to lower target budget consumption values.

[0058] The server can then calculate the deviation between the current budget consumption data and the target budget consumption value, and determine the bid adjustment factor.

[0059] Specifically, the server can determine the bid adjustment coefficient based on the aforementioned deviation, combined with the historical cumulative amount of the deviation and / or the trend of the deviation. The historical cumulative amount of the deviation reflects the cumulative budget execution deviation over a period of time, while the trend of the deviation reflects the direction and rate of increase of the budget deviation. The server can determine the bid adjustment coefficient based on a combination of proportional, integral, and derivative terms, or it can use other calculation methods that can utilize current deviation, historical cumulative amount, and trend for closed-loop control.

[0060] For example, if the current budget consumption data is significantly lower than the target budget consumption value, and this deviation continues to accumulate over multiple consecutive statistical periods, the server can increase the bid adjustment coefficient based on the historical cumulative amount of the deviation to enhance the competitiveness of budget-controlled creatives and accelerate budget consumption.

[0061] When the current budget consumption data is higher than the target budget consumption value, and the deviation shows a continuous widening trend in adjacent statistical periods, the server can reduce the bid adjustment coefficient based on the trend of the deviation to suppress further overspending.

[0062] When there is a significant deviation between the current budget consumption data and the target budget consumption value, and this deviation has both historical accumulation and a trend of continuous increase or decrease, the server can combine the historical accumulation of the deviation and the trend of the deviation to jointly determine the bid adjustment coefficient, so that the budget control process takes into account both response speed and execution stability.

[0063] In addition, when the current budget consumption data is close to the target budget consumption value, or when the historical cumulative amount of deviation and the trend of deviation change are both within a small range, the server can keep the bid adjustment factor near the benchmark range to maintain smooth budget execution.

[0064] Optionally, the server can also set upper and lower limits for the bid adjustment coefficient to avoid system oscillations caused by excessive budget control. For example, the bid adjustment coefficient can be limited to a preset minimum and a preset maximum value. When the calculated result exceeds the maximum value, the maximum value is used; when the calculated result is lower than the minimum value, the minimum value is used. These upper and lower limits can be determined based on business experience, experimental results, or the budget sensitivity of different booths.

[0065] Before determining the bid adjustment factor, the server can also obtain the predicted campaign performance data corresponding to the target creatives determined based on different processing links, compare the predicted campaign performance data corresponding to each processing link, and determine the target processing link from among the processing links based on the comparison results. The determination strategy and / or calculation parameters for the target creatives differ for each processing link. This processing method can be used for experimental comparisons of old and new links, multi-strategy switching, or experimental traffic splitting scenarios. For example, the first processing link can use the first set of budget control parameters, and the second processing link can use the second set of budget control parameters. After comparing the campaign performance predicted by the two links, the server selects the target processing link that is more suitable for the current experimental objective, and continues to execute the bid adjustment factor determination process under this target processing link.

[0066] The determination strategies corresponding to different processing links can include the filtering order of candidate creatives, the sorting method of target creatives, the determination method of dynamic recommendation score, or the selection method of fallback strategy. The calculation parameters corresponding to different processing links can include: parameters used when calculating the bid adjustment coefficient, such as deviation weight coefficient, historical cumulative amount coefficient, and change trend coefficient; parameters used when calculating the dynamic recommendation score, such as matching score weight, creative bid weight, and future budget consumption data weight; parameters set during the screening process, such as screening threshold, fatigue threshold, and budget limit parameter; and default parameters used in the fallback processing, such as default recommendation score, default bid parameter, or default sorting parameter.

[0067] For example, in a comparison scenario between old and new versions, the old version's processing link can use the default matching score weight, default filtering threshold, and fixed bid adjustment parameters, while the new version's processing link can use updated matching score weight, dynamically changing filtering thresholds, and reset deviation weight coefficients, historical cumulative coefficients, and / or trend coefficients. The server can compare the predicted click-through rate, conversion rate, or overall revenue for the target creative determined by different processing links and select the processing link with the better predicted campaign performance as the target processing link.

[0068] In practical applications, the aforementioned bid adjustment coefficient can be calculated based on the Proportional-Integral-Derivative (PID) control formula. Specifically, the server can calculate the bid adjustment coefficient according to the current deviation between the current budget consumption data and the target budget consumption value, the historical cumulative amount of the deviation, and the trend of the deviation, based on preset proportional, integral, and derivative coefficients.

[0069] In some implementations, the bid adjustment factor can also be obtained through a pre-trained prediction or control model, such as a machine learning model, a deep learning model, or a reinforcement learning model. The input to this model may include current budget consumption data, target budget consumption value, and historical bid data, and the model output is the corresponding bid adjustment factor.

[0070] The training samples for the above model can be derived from historical ad delivery logs, budget execution logs, user behavior logs, or experimental data. The server can train the model based on budget adjustment results under different budget execution states in historical samples, enabling the model to learn the mapping relationship between budget execution states and bid adjustment coefficients. S206: Determine the dynamic recommendation score corresponding to the budget-controlled ad based on the matching score and the ad bid adjusted based on the bid adjustment coefficient.

[0071] In other optional embodiments of this specification, a lookup table method based on rule thresholds or a linear regression mapping method can also be used to efficiently determine the coefficient, in order to adapt to scenarios with different computing resource constraints or real-time requirements.

[0072] Specifically, the server can pre-build a mapping table between "budget deviation range" and "bid adjustment coefficient". This mapping table can be configured based on historical campaign data or business experience. For example, the deviation between the current budget consumption data and the target budget consumption value can be divided into multiple levels: when the deviation is in the first range (e.g., actual consumption is more than 20% lower than the target value), a larger preset coefficient (e.g., 1.2) is directly read from the mapping table as the bid adjustment coefficient to improve the competitiveness of the creative; when the deviation is in the second range (e.g., the deviation between actual consumption and the target value is within ±5%), a baseline coefficient (e.g., 1.0) is read; when the deviation is in the third range (e.g., actual consumption is more than 20% higher than the target value), a smaller preset coefficient (e.g., 0.8) is read to suppress consumption.

[0073] The server can establish a linear functional relationship between the budget deviation value and the bid adjustment factor. In this mode, the bid adjustment factor transitions linearly and smoothly with changes in budget deviation: the larger the deviation (i.e., the more the budget consumption lags behind), the coefficient increases linearly; the smaller the deviation or the negative value (i.e., the budget consumption is ahead of schedule), the coefficient decreases linearly.

[0074] Those skilled in the art should understand that, regardless of whether PID control, time series forecasting, rule lookup, or linear mapping is used, the essence is to dynamically generate adjustment factors based on the "difference between the current budget consumption state and the target state," and all fall within the technical scope of "determining the bid adjustment coefficient based on the deviation" protected in this specification.

[0075] The server can first adjust the creative bid of the budget-controlled creative based on the bid adjustment factor to obtain the adjusted creative bid. The creative bid can be the base bid value of the creative's original configuration, or the initial bid value of the creative in a certain ad unit. By introducing the bid adjustment factor, the server can dynamically change the bidding ability of the budget-controlled creative in the current request based on the budget execution status.

[0076] The server can then determine a dynamic recommendation score based on the matching score and the adjusted creative bid.

[0077] Specifically, the server can determine the budget score corresponding to the budget-controlled creative based on the adjusted creative bid. This budget score is positively correlated with the adjusted creative bid; that is, the higher the adjusted creative bid, the higher the corresponding budget score. In practice, the server can input the adjusted creative bid into a preset budget score calculation model, or convert the adjusted creative bid into a budget score based on preset mapping rules.

[0078] After obtaining the budget score, the server can determine the dynamic recommendation score corresponding to the budget-controlled creative based on the matching score and the budget score.

[0079] Furthermore, the server can incorporate a future budget consumption score into the process of determining the dynamic recommendation score. Specifically, the server can predict future budget consumption data based on current budget consumption data and bid adjustment coefficients, and determine the future budget consumption score based on this future budget consumption data.

[0080] The future budget consumption data is used to characterize the budget consumption of budget control ideas over a future time period. Specifically, it can reflect at least one of the following: future budget consumption trend, future budget consumption speed, future budget consumption rhythm, or future budget deficit level. The server can determine the future budget consumption score based on this data, thus converting the future budget consumption into a quantitative indicator that can participate in the calculation of dynamic recommendation scores.

[0081] The future budget consumption score can be positively correlated with the degree of future budget deficit, that is, the larger the future budget deficit, the higher the future budget consumption score; or, the future budget consumption score can be negatively correlated with the risk of budget overspending, that is, the higher the risk of budget overspending, the lower the future budget consumption score.

[0082] In some implementations, future budget consumption data can be predicted based on a time series forecasting model. For example, the server can input at least one of the following into a time series forecasting model: historical budget consumption data, historical traffic data, historical exposure data, click data, conversion data, and time features, to obtain budget consumption prediction results for one or more future time slices, which can then be used as future budget consumption data. The time series forecasting model can be a Long Short-Term Memory (LSTM) network, a Transformer model, or other models suitable for time series forecasting.

[0083] After obtaining the future budget consumption score, the server can determine the dynamic recommendation score corresponding to the budget-controlled creative based on the matching score, budget score, and future budget consumption score. For example, the server can multiply the matching score by a first weight, the budget score by a second weight, and the future budget consumption score by a third weight, and then perform a weighted sum to obtain the dynamic recommendation score.

[0084] Furthermore, the first, second, and third weights can be determined based on booth type, advertising objectives, experimental parameters, or business strategy configuration. For example, if future budget consumption data indicates that the current booth will still have a significant budget shortfall in the subsequent period, the server can determine a higher future budget consumption score or appropriately increase the weight corresponding to the future budget consumption score to enhance the exposure opportunity of budget-controlled creatives; if future budget consumption data indicates that the budget is at risk of being overdrawn in the subsequent period, the server can determine a lower future budget consumption score or appropriately reduce the weight corresponding to the future budget consumption score to avoid subsequent budget overdraft.

[0085] In some embodiments, the server may also normalize the matching score, budget score, and future budget consumption score separately before merging them to avoid uneven impact of different units on the dynamic recommendation score.

[0086] S208: Determine the target creative based on the dynamic recommendation score corresponding to each budget-controlled creative and deliver it to the user.

[0087] In this specification, candidate creatives may include both budget-controlled and non-budget-controlled creatives. In this case, the server can first collect the dynamic recommendation score for each budget-controlled creative and the preset recommendation score for each non-budget-controlled creative, and map both to the same scoring space to form comparable scoring results for each candidate creative. Subsequently, the server can determine the target creative based on these scoring results.

[0088] In one alternative embodiment, the server may first determine a filtering threshold based on at least one of preset calculation parameters, target budget consumption status, and advertising campaign objectives. Advertising campaign objectives may include click objectives, conversion objectives, impression objectives, dwell objectives, or overall performance objectives. The filtering threshold may be a fixed value or a value dynamically determined based on the current budget status, ad space competition intensity, and experimental configuration.

[0089] Once the filtering threshold is obtained, the server can filter candidate creatives based on it. For example, the server can filter out candidate creatives with scores below the filtering threshold, retaining only those with scores higher than or equal to the threshold. The server then sorts the filtered candidate creatives and determines the target creative based on the sorting results. Sorting methods can include direct sorting by score from highest to lowest, binning by score and then performing a secondary sort within each bin, or multi-dimensional sorting based on scores combined with other sorting factors.

[0090] In one implementation, the server can directly identify the candidate creative with the highest score as the target creative. In another implementation, the server can select the top K candidate creatives to form a target creative set, and then determine the final target creative from the target creative set according to the deduplication rule, the carousel rule, or the booth display capacity. In another implementation, for booths that support the simultaneous display of multiple creative ideas, the server can select multiple target creative ideas according to the sorting results and display them to the user in a preset order.

[0091] In practical applications, if the number of candidate creatives remaining after fatigue filtering and threshold screening is small, the server can also trigger fallback logic. Fallback logic may include replenishing creatives from the default creative pool, calling historical high-quality creatives, directly using non-budget-controlled creatives, or executing the default recommendation strategy to ensure that the ad placement can complete the ad display normally.

[0092] Of course, the candidate creatives may also include only budget-controlled creatives. In this case, the server can determine the target creatives based on the dynamic recommendation scores corresponding to each budget-controlled creative and deliver them to the user.

[0093] In one performance optimization-related embodiment, at least some steps in determining the target creative idea can be executed in parallel. For example, at least some steps in obtaining the matching score, obtaining budget-related data, and determining the fatigue score can be executed in parallel. For each parallel execution step, if the execution time of the step exceeds a preset time threshold, the server can degrade the step to serial execution; for each serial execution step, if the step fails or the execution time exceeds the preset time threshold, the server can use the corresponding default result or historical result as the execution result of the step, and determine the target creative idea based on the execution results of each step. Through this mechanism combining parallelism and degradation, the stability of the system in a high-concurrency environment can be improved while ensuring the main link response latency.

[0094] For example, the server can allocate independent threads for tasks such as matching score retrieval, budget status retrieval, and fatigue calculation, and wait for the parallel tasks to return within a preset total time window. If a parallel task fails to complete within the time limit, the server can switch the task to serial compensation execution; if the serial compensation execution also fails or times out again, the server can directly use the default or historical result corresponding to that task. For example, in the case of matching score retrieval failure, the server can use a historical matching score or a preset default matching score; in the case of budget status retrieval failure, the server can use the historical result corresponding to the most recent valid budget status; in the case of fatigue calculation failure, the server can use the default no-fatigue result or the most recent fatigue result. This processing method can avoid the entire ad delivery request timeout failure due to a single task anomaly.

[0095] To facilitate understanding, an exemplary advertising delivery process is given below.

[0096] Suppose a user triggers an ad display request on the homepage of a certain business. The server first determines the ad placement information for a specific ad slot on the homepage and retrieves ten candidate creatives from the creative pool corresponding to that slot, six of which are budget-controlled creatives and four are non-budget-controlled creatives. The server further retrieves the current budget consumption data for that slot and finds that the current budget consumption is lower than the target budget consumption for the current time point. Subsequently, based on the creative identifiers of each budget-controlled creative and the user's identifier, the server retrieves the corresponding matching scores from the first-level and second-level caches. For individual creatives that do not match the target score, the server calls the recommendation service in real time to obtain the matching score. Next, the server calculates a bid adjustment coefficient greater than the baseline value based on the current budget deviation, historical cumulative deviation, and deviation trend, and uses this coefficient to adjust the bids of the six budget-controlled creatives. Based on the adjusted creative bids and matching scores, the server obtains the dynamic recommendation scores for each of the six budget-controlled creatives. For the four non-budget-controlled creatives, the server directly reads the corresponding preset recommendation scores. Next, the server performs fatigue filtering and threshold filtering on all candidate creatives, sorts the remaining creatives by score, and finally determines the top-ranked candidate creative as the target creative, which is then delivered to that user. This process can appropriately increase the visibility of budget-controlled creatives when current budget execution is slow, thereby improving the budget execution rhythm.

[0097] Furthermore, this specification provides an overall flowchart for advertising placement, such as... Figure 3 As shown.

[0098] Upon receiving an ad delivery request, the server first retrieves candidate creatives and then determines whether each candidate creative falls under the budget control category. For budget control creatives, the server obtains the matching score, calculates the bid adjustment coefficient, performs fatigue filtering, and then calculates the dynamic recommendation score. These three steps can be executed in parallel.

[0099] For creatives not subject to budget control, a preset recommendation score can be directly obtained. After determining the recommendation scores of all candidate creatives, the server sorts the candidate creatives and selects the target creative for delivery.

[0100] The process of determining the recommendation score for each candidate idea can be executed in parallel.

[0101] Furthermore, this specification also provides a logical architecture diagram for ad delivery, such as... Figure 4 As shown.

[0102] The system comprises several layers: a data acquisition and preprocessing layer, a core business processing layer, and a management layer. The data acquisition and preprocessing layer is used to acquire and organize the basic data required for ad delivery, including budget consumption data, historical budget allocation data, user behavior data, and cached data, providing data support for subsequent business processing. The core business processing layer is used to execute core processing logic such as budget smoothing control, dynamic bidding calculation, creative quality evaluation, and experimental traffic allocation and effect evaluation. The strategy execution and optimization layer is used to orchestrate and execute various processing strategies and supports functions such as parallel collaboration, anomaly degradation, and real-time alarms to improve system stability and processing efficiency. The service interface layer provides capabilities such as external interface calls, remote service calls, asynchronous message communication, and cache synchronization to support internal and external interactions. The management and monitoring layer implements functions such as parameter configuration, performance monitoring, experimental data display, and operational effect analysis to facilitate unified management of system operation status and delivery performance.

[0103] Figure 5 This is a schematic structural diagram of a device provided in an exemplary embodiment. For example... Figure 5 As shown, device 500 mainly consists of a communication interface 502, a user interface 504, a processor 506, and a data storage 508. These components are interconnected and communicate with each other via a system bus, network, or other connection mechanism 510. The communication interface 502 enables device 500 to communicate with other devices, access networks, and transmission networks via analog or digital modulation. For example, the communication interface 502 may include a chipset and antenna for wireless communication with a radio access network or access point. Furthermore, the communication interface 502 can be a wired interface such as Ethernet, Token Ring, or a USB port, or a wireless interface such as Wi-Fi, Bluetooth, Global Positioning System (GPS), or a wide-area wireless interface (e.g., WiMAX or LTE). Of course, the communication interface 502 can also support other forms of physical layer interfaces and standard or proprietary communication protocols. The communication interface 502 may also include multiple physical communication interfaces, such as Wi-Fi interfaces, Bluetooth interfaces, and wide-area wireless interfaces.

[0104] User interface 504 includes receiving user input and providing output to the user. Therefore, user interface 504 may include input components such as a keypad, keyboard, touch-sensitive or presence-sensitive panel, computer mouse, trackball, joystick, microphone, still camera, and video camera, and output components such as a display screen (which may be combined with a touch-sensitive panel), CRT, LCD, LED, display using DLP technology, printer, and other similar devices known or developed in the future. User interface 504 may also generate auditory output via speakers, speaker jacks, audio output ports, audio output devices, headphones, and other similar devices known or developed in the future. In some embodiments, user interface 504 may include software, circuitry, or other forms of logic capable of transmitting and receiving data from external user input / output devices. Additionally or alternatively, device 500 may support remote access from other devices via communication interface 502 or another physical interface (not shown). User interface 504 may be configured to receive user input, the position and movement of which may be indicated by indicators or cursors described herein. User interface 504 may also be configured as a display device for rendering or displaying text fragments.

[0105] Processor 506 may contain one or more general-purpose processors and / or special-purpose processors.

[0106] Data storage 508 may include one or more volatile and / or non-volatile storage components and may be integrated wholly or partially with processor 506. Data storage 508 may include removable and non-removable components.

[0107] Processor 506 is capable of executing program instructions 518 (e.g., compiled or uncompiled program logic and / or machine code) stored in data storage 508 to perform the various functions described herein. Data storage 508 may contain a non-transitory computer-readable medium on which program instructions are stored, which, when executed by device 500, enable device 500 to perform any methods, processes, or functions disclosed in this specification and / or the accompanying drawings. Execution of program instructions 518 by processor 506 may result in processor 506 using data 512.

[0108] For example, program instructions 518 may include an operating system 522 (e.g., an operating system kernel, device drivers, and / or other modules) installed on device 500 and one or more applications 520 (e.g., a browser, social application, or game application). Similarly, data 512 may include operating system data 516 and application data 514. Operating system data 516 is primarily accessible to the operating system 522, while application data 514 is primarily accessible to one or more applications 520. Application data 514 may reside in a file system visible or hidden from the user of device 500.

[0109] Application 520 can communicate with operating system 522 through one or more application programming interfaces (APIs). These APIs help application 520 read and / or write application data 514, transmit or receive information via communication interface 502, receive or display information on user interface 504, etc.

[0110] In some terminology, application 520 may be simply referred to as "app". Furthermore, application 520 can be downloaded to device 500 through one or more online app stores or app markets. However, applications can also be installed on device 500 in other ways, such as through a web browser or a physical interface on device 500 (e.g., a USB port).

[0111] Please refer to Figure 6 Advertising placement devices can be applied to, for example Figure 5 The device shown is used to implement the technical solution described in this specification. The advertising delivery device may include: The acquisition unit 600 acquires the booth information of the current advertising page and the candidate creatives corresponding to the booth information based on the advertising trigger operation performed by the user; wherein, the candidate creatives include budget-controlled creatives subject to budget constraints. The determining unit 602 determines the current budget consumption data of the booth information, and for each budget control creative, obtains the matching score between the budget control creative and the user; The calculation unit 604 determines the bid adjustment coefficient based on the deviation between the current budget consumption data and the target budget consumption status; The adjustment unit 606 determines the dynamic recommendation score corresponding to the budget-controlled creative based on the matching score and the creative bid adjusted based on the bid adjustment coefficient. The delivery unit 608 determines the target creative based on the dynamic recommendation score corresponding to each budget-controlled creative and delivers it to the user.

[0112] Optionally, the acquisition unit 600 is further configured to acquire the user's activity status data; if the user is determined to be a highly active user based on the activity status data, then no advertising is delivered to the user or the non-budget-controlled creative is delivered to the user; if the user is determined to be a low-active user based on the activity status data, then a dynamic recommendation score determination process is performed for each candidate creative.

[0113] Optionally, the determining unit 602 is specifically used to determine a cache index identifier based on the creative identifier of the budget control creative and the user identifier of the user, and retrieve the corresponding matching score in a preset first-level cache based on the cache index identifier; if the corresponding matching score is not found in the first-level cache, then the matching score corresponding to the cache index identifier is retrieved in the second-level cache; wherein, the first-level cache includes local memory, the second-level cache includes a distributed cache, and the first-level cache stores the matching scores corresponding to each index identifier with an access frequency higher than a preset frequency.

[0114] Optionally, the determining unit 602 is further configured to, when a preset detection time is reached, obtain the cache timestamps corresponding to each cache index identifier in the first-level cache and the second-level cache; calculate the time interval between the preset detection time and the cache timestamps corresponding to each cache index identifier; and update the matching score corresponding to any cache index identifier if the time interval corresponding to any cache index identifier exceeds a preset time threshold.

[0115] Optionally, the calculation unit 604 is specifically used to determine the bid adjustment coefficient based on the deviation between the current budget consumption data and the target budget consumption status.

[0116] Optionally, the determining unit 602 is further configured to: acquire the predicted delivery effect data corresponding to the target creative determined based on different processing links; compare the predicted delivery effect data corresponding to each processing link; and determine the target processing link from each processing link based on the comparison results; wherein the determination strategy and / or calculation parameters of the target creative corresponding to each processing link are different.

[0117] Optionally, the target budget consumption status includes the budget ceiling and the target budget consumption value corresponding to at least one time point; The calculation unit 604 is specifically used to: adjust the basic budget upper limit of the booth information according to historical traffic trends to obtain the adjusted budget upper limit; determine the target budget consumption value for the current time node according to the adjusted budget upper limit; and determine the bid adjustment coefficient according to the deviation between the budget consumption data for the current time node and the target budget consumption value, combined with the historical cumulative amount of the deviation and the changing trend of the deviation.

[0118] Optionally, the adjustment unit 606 is specifically used to predict future budget consumption data based on the current budget consumption data and the bid adjustment coefficient; and to determine the dynamic recommendation score corresponding to the budget-controlled creative based on the matching score, the adjusted creative bid, the future budget consumption data, and their respective weights.

[0119] Optionally, the delivery unit 608 is specifically used to: determine a screening threshold based on at least one of preset calculation parameters, target budget consumption status, and advertising delivery objectives; screen each candidate creative according to the screening threshold; and sort the screened candidate creatives to determine the target creative based on the sorting results.

[0120] Optionally, the delivery unit 608 is further configured to determine the target fatigue score corresponding to each candidate creative; and filter out candidate creatives whose corresponding target fatigue score is greater than a preset score from each candidate creative.

[0121] Optionally, the delivery unit 608 is further configured to: determine a business dimension fatigue score based on the historical recommendation data of the user across the entire business chain; determine a booth dimension fatigue score based on the historical recommendation data of the user on the delivery page; determine a creative dimension fatigue score for each candidate creative based on the historical recommendation data of each candidate creative for the user; and for each candidate creative, determine a target fatigue score for that candidate creative based on the business dimension fatigue score, the booth dimension fatigue score, and the creative dimension fatigue score corresponding to that candidate creative.

[0122] Optionally, the candidate ideas also include: non-budget-controlled ideas that are not subject to budget constraints; The delivery unit 608 is specifically used to determine the target creative and deliver it to the user based on the dynamic recommendation score corresponding to each budget-controlled creative and the preset recommendation score corresponding to each non-budget-controlled creative.

[0123] For ease of description, the above devices are described by dividing them into various modules or units based on their functions. Of course, when implementing one or more of these specifications, the functions of each module or unit can be implemented in one or more software and / or hardware components, or a module that performs the same function can be implemented by a combination of multiple sub-modules or sub-units. The device embodiments described above are merely illustrative. For example, the division of units is only a logical functional division; in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed.

[0124] Based on the same concept as the methods described above, this specification also provides an electronic device, including: a processor; a memory for storing processor-executable instructions; wherein the processor performs the steps of the method as described in any of the above embodiments by executing the executable instructions.

[0125] Based on the same concept as the methods described above, this specification also provides a computer-readable storage medium having computer instructions stored thereon that, when executed by a processor, implement the steps of the methods as described in any of the above embodiments.

[0126] Based on the same concept as the methods described above, this specification also provides a computer program product, including a computer program / instructions that, when executed by a processor, implement the steps of the methods as described in any of the above embodiments.

[0127] What those skilled in the art will understand is: In this specification, the terms "comprising," "including," or any other variations thereof are intended to cover a non-exclusive inclusion, such that a process, method, product, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, product, or apparatus. Without further limitation, the presence of additional identical or equivalent elements in a process, method, product, or apparatus that includes said elements is not excluded.

[0128] In this specification, “a,” “an,” and “the” do not specifically refer to the singular, but may also include the plural.

[0129] In this specification, ordinal numbers such as "first," "second," etc., do not necessarily indicate order; they are often used to distinguish between objects. For example, "first server" and "second server" usually refer to two servers. To differentiate between these two servers, they are described as "first server" and "second server." Of course, sometimes these two servers may be the same server.

[0130] In this specification, unless explicitly stated otherwise, "receiving and sending data" does not necessarily mean direct receiving and sending; it can also mean indirect receiving and sending. For example, A receiving data sent by B can be understood as A directly receiving the data sent by B, or it can be understood as A indirectly receiving the data sent by B through other entities such as C. Similarly, B sending data to A can be understood as B sending the data directly to A, or it can be understood as B indirectly sending the data to A through other entities such as C. Here, C can be one entity, or it can be two or more entities.

[0131] In this specification, unless explicitly stated otherwise, the relationships between structures can be direct or indirect. For example, when describing "A is connected to B," unless it is explicitly stated that A and B are directly connected, it should be understood that A can be directly connected to B or indirectly connected to B. Similarly, when describing "A is on top of B," unless it is explicitly stated that A is directly above B (AB is adjacent and A is above B), it should be understood that A can be directly above B or indirectly above B (AB is separated by other elements, and A is above B). And so on.

[0132] This specification uses specific terms to describe embodiments thereof. Terms such as "an embodiment," "one embodiment," and / or "some embodiments" refer to a particular feature, structure, or characteristic associated with at least one embodiment of this specification. Therefore, it should be emphasized and noted that references to "an embodiment," "one embodiment," or "an alternative embodiment" in different locations throughout this specification do not necessarily refer to the same embodiment. Furthermore, those skilled in the art can combine and integrate the different embodiments or examples described herein, as well as the features of those different embodiments or examples, without contradiction.

[0133] Although one or more embodiments of this specification provide method steps as described in the embodiments or flowcharts, it is understood that the order of steps listed in the embodiments or flowcharts is only one of many possible execution orders and does not represent the only execution order. Therefore, when the claims involve method steps, any changes or adjustments to the order of such steps, or the parallelism between steps, are also within the scope of protection of the claims.

Claims

1. An advertising placement method, comprising: Based on the user's execution of the campaign trigger operation, obtain the booth information of the current campaign page and the candidate creatives corresponding to the booth information; wherein, the candidate creatives include budget-controlled creatives subject to budget constraints; Determine the current budget consumption data of the booth information, and for each budget control creative, obtain the matching score between the budget control creative and the user; The bid adjustment factor is determined based on the deviation between the current budget consumption data and the target budget consumption status; Based on the matching score and the creative bid adjusted based on the bid adjustment coefficient, the dynamic recommendation score corresponding to the budget-controlled creative is determined; Target creatives are determined based on the dynamic recommendation scores corresponding to each budget-controlled creative, and then delivered to the users.

2. The method of claim 1, further comprising: Obtain the user's activity status data; If the user is determined to be a highly active user based on the activity status data, then no ads will be delivered to the user or no non-budget-controlled creatives will be delivered to the user. If the user is determined to be a low-activity user based on the activity status data, then a dynamic recommendation score determination process is performed for each candidate creative.

3. The method as described in claim 1, obtaining the matching score between the budget control idea and the user, specifically includes: Based on the creative identifier of the budget control creative and the user identifier of the user, a cache index identifier is determined, and the corresponding matching score is retrieved from the preset first-level cache based on the cache index identifier; If the corresponding matching score is not found in the first-level cache, the matching score corresponding to the cache index identifier is retrieved in the second-level cache. The first-level cache includes local memory, the second-level cache includes a distributed cache, and the first-level cache stores the matching score corresponding to each index identifier whose access frequency is higher than a preset frequency.

4. The method as described in claim 3, characterized in that, The method further includes: When the preset detection time is reached, obtain the cache timestamp corresponding to each cache index identifier in the first-level cache and the second-level cache; Calculate the time interval between the preset detection time and the cache timestamp corresponding to each cache index identifier; If the time interval corresponding to any cached index identifier exceeds the preset time threshold, the matching score corresponding to that cached index identifier will be updated.

5. The method as described in claim 1, wherein determining the bid adjustment coefficient specifically includes: Based on the deviation between the current budget consumption data and the target budget consumption status, and combined with the historical cumulative amount and / or trend of the deviation, the bid adjustment coefficient is determined.

6. The method of claim 1, wherein, Before determining the bid adjustment factor, the method further includes: Obtain the predicted delivery performance data for the target creatives determined based on different processing links; The predicted delivery performance data corresponding to each of the processing links are compared, and the target processing link is determined from each of the processing links based on the comparison results. The determination strategies and / or calculation parameters for the target creative are different for each of the processing links.

7. The advertising delivery method as described in claim 1, wherein the target budget consumption status includes a budget cap and a target budget consumption value corresponding to at least one time point; Determine the bid adjustment factor, specifically including: The basic budget ceiling for the booth information is adjusted based on historical traffic trends to obtain the adjusted budget ceiling. The target budget consumption value for the current time node is determined based on the adjusted budget ceiling. The bid adjustment coefficient is determined based on the deviation between the current budget consumption data and the target budget consumption value, combined with the historical cumulative amount of the deviation and the trend of the deviation.

8. The method as described in claim 1, determining the dynamic recommendation score corresponding to the budget-controlled creative, specifically includes: Based on the current budget consumption data and the bid adjustment coefficient, predict future budget consumption data; Based on the matching score, the adjusted creative bid, the future budget consumption data, and their respective weights, the dynamic recommendation score corresponding to the budget-controlled creative is determined.

9. The method as described in claim 1, wherein the target creative is determined based on the dynamic recommendation score corresponding to each budget-controlled creative, specifically includes: The selection threshold is determined based on at least one of the following: preset calculation parameters, target budget consumption status, and advertising campaign objectives; Each candidate idea is filtered according to the aforementioned screening threshold; The selected candidate ideas are sorted to determine the target idea based on the sorting results.

10. The advertising placement method as described in claim 1, further comprising, before determining the target creative: Determine the target fatigue score for each candidate idea; Filter out candidate ideas whose corresponding target fatigue score is greater than the preset score from each candidate idea.

11. The method as described in claim 10, determining the fatigue score corresponding to each candidate idea, specifically includes: Based on the user's historical recommendation history across the entire business process, a fatigue score for the business dimension is determined. Based on the user's historical recommendations on the advertising page, determine the fatigue score for the advertising space. Based on the historical recommendation history of each candidate creative for the user, determine the fatigue score of each candidate creative for the creative dimension; For each candidate creative, the target fatigue score is determined based on the fatigue score of the business dimension, the fatigue score of the booth dimension, and the fatigue score of the creative dimension corresponding to the candidate creative.

12. The method of claim 1, wherein the candidate idea further comprises: Unconstrained, non-budget-controlled creative ideas; Identifying target creative content and delivering it to the specified users includes: Based on the dynamic recommendation score corresponding to each budget-controlled creative and the preset recommendation score corresponding to each non-budget-controlled creative, the target creative is determined and delivered to the user.

13. An advertising delivery device, comprising: The acquisition unit acquires the booth information of the current ad placement page and the candidate creatives corresponding to the booth information based on the ad placement trigger operation performed by the user; wherein, the candidate creatives include budget-controlled creatives subject to budget constraints. The unit determines the current budget consumption data of the booth information and, for each budget control creative, obtains the matching score between the budget control creative and the user. The calculation unit determines the bid adjustment coefficient based on the deviation between the current budget consumption data and the target budget consumption status; The adjustment unit determines the dynamic recommendation score corresponding to the budget-controlled creative based on the matching score and the creative bid adjusted based on the bid adjustment coefficient. The delivery unit determines the target creative based on the dynamic recommendation score corresponding to each budget-controlled creative and delivers it to the user.

14. An electronic device comprising: processor; A memory for storing processor-executable instructions; wherein the processor implements the steps of the method as described in any one of claims 1-12 by executing the executable instructions.

15. A computer program product comprising a computer program / instructions that, when executed by a processor, implement the steps of the method as claimed in any one of claims 1-12.

Images