Content pushing method and device, computer device and storage medium

Abstract

The application relates to a content pushing method and device, computer equipment and a storage medium. The method comprises the following steps: determining candidate resource consumption factors of candidate pushing contents on at least two business targets for a target object, obtaining each candidate resource consumption factor, determining a target resource consumption factor corresponding to the candidate pushing contents based on the candidate resource consumption factors, obtaining a target resource consumption amount of the candidate pushing contents for the target object based on the target resource consumption factor of the candidate pushing contents and an expected resource consumption amount of the candidate pushing contents for the target object, and performing content pushing processing on the target object based on the target resource consumption amount. The method can reduce the influence on the resource consumption amount. The content pushing method provided by the application can be applied to the field of advertisement putting.

Description

Technical Field

[0001] This application relates to the field of computer technology, and in particular to a content delivery method, apparatus, computer device, and storage medium. Background Technology

[0002] With the development of computer and internet technologies, more and more content is being pushed online, such as advertising. For example, advertisers pay traffic owners to push their ads through the traffic owners' media platforms.

[0003] In traditional methods, content delivery typically involves adjusting the resource consumption of multiple pieces of content based on a strategy, and then pushing content according to the adjusted resource consumption. Here, the strategy refers to the business objective, such as "increasing paid subscriptions for game ads." Resource consumption refers to the amount of resources required to push content.

[0004] However, the method of adjusting resource consumption using strategies has the problem of a large difference between the adjusted resource consumption and the expected resource consumption, resulting in a significant impact on resource consumption. Summary of the Invention

[0005] Therefore, it is necessary to provide a content delivery method, apparatus, computer device, computer-readable storage medium, and computer program product that can reduce the impact on resource consumption in order to address the above-mentioned technical problems.

[0006] On one hand, this application provides a content push method. The method includes: for candidate push content for a target object, determining the resource consumption factors corresponding to the candidate push content on at least two business objectives respectively, to obtain each candidate resource consumption factor; determining the target resource consumption factor corresponding to the candidate push content based on each candidate resource consumption factor; obtaining the target resource consumption amount of the candidate push content for the target object based on the target resource consumption factor corresponding to the candidate push content and the expected resource consumption amount of the candidate push content for the target object; the expected resource consumption amount is the expected resource consumption amount when the push object of the candidate push content is the target object; and performing content push processing on the target object based on the target resource consumption amount.

[0007] On the other hand, this application also provides a content push device. The device includes: a candidate resource consumption factor obtaining module, configured to determine, for candidate push content for a target object, the resource consumption factors corresponding to the candidate push content on at least two business objectives, thereby obtaining each candidate resource consumption factor; a target resource consumption factor determining module, configured to determine the target resource consumption factor corresponding to the candidate push content based on each candidate resource consumption factor; a target resource consumption amount obtaining module, configured to obtain the target resource consumption amount of the candidate push content for the target object based on the target resource consumption factor corresponding to the candidate push content and the expected resource consumption amount of the candidate push content for the target object; the expected resource consumption amount is the expected resource consumption amount when the push object of the candidate push content is the target object; and a content push processing module, configured to perform content push processing on the target object based on the target resource consumption amount.

[0008] In some embodiments, the target resource consumption factor determination module is further configured to determine the data distribution information of each candidate resource consumption factor; and based on the data distribution information, select the target resource consumption factor corresponding to the candidate push content from each candidate resource consumption factor.

[0009] In some embodiments, the target resource consumption factor determination module is further configured to determine the number of candidate resource consumption factors greater than a preset value among the candidate resource consumption factors, to obtain a first number; determine the number of candidate resource consumption factors less than the preset value among the candidate resource consumption factors, to obtain a second number; and select the target resource consumption factor corresponding to the candidate push content from among the candidate resource consumption factors based on the first number and the second number.

[0010] In some embodiments, the target resource consumption factor determination module is further configured to determine a factor screening strategy based on the first quantity and the second quantity when the first quantity is different from the second quantity; and select the target resource consumption factor corresponding to the candidate push content from the candidate resource consumption factors based on the factor screening strategy.

[0011] In some embodiments, the candidate resource consumption factor obtaining module is further configured to determine, for the target object, the initial resource consumption factors corresponding to the candidate push content on the at least two business objectives respectively; for each of the at least two business objectives, obtain a first factor adjustment coefficient for the candidate push content; and adjust the initial resource consumption factor of the candidate push content on the business objective using the first factor adjustment coefficient for the candidate push content to obtain the candidate resource consumption factor of the candidate push content on the business objective.

[0012] In some embodiments, there are multiple target objects, and each target object corresponds to multiple candidate push content; the apparatus further includes: a target push content selection module, configured to, for each target object, select target push content corresponding to the target object from the multiple candidate push content corresponding to the target object based on the target resource consumption of the target object; a content push module, configured to, for each target object, push the target push content corresponding to the target object to the target object's terminal; a coefficient update module, configured to, based on multiple historical push records within a first historical time period, update the first factor adjustment coefficients of the at least two business objectives for the candidate push content; and return to the step of adjusting the initial resource consumption factor of the candidate push content on the business objective using the first factor adjustment coefficients of the business objective for the candidate push content, to obtain the candidate resource consumption factor of the candidate push content on the business objective.

[0013] In some embodiments, the coefficient update module is further configured to, for each of the at least two business objectives, select from a plurality of historical push records a historical push record that records the identifier of the candidate push content and the identifier of the business objective, to obtain a reference push record corresponding to the business objective; determine a reference object corresponding to the business objective based on the object identifier recorded in each reference push record corresponding to the business objective; determine the initial resource consumption factor of the candidate push content on the business objective for the reference object; and update the first factor adjustment coefficient of the business objective for the candidate push content based on the initial resource consumption factor.

[0014] In some embodiments, the coefficient update module is further configured to obtain the first resource consumption of the reference object corresponding to the business objective for the candidate push content; the first resource consumption refers to the expected resource consumption under a preset resource consumption strategy; and update the first factor adjustment coefficient of the business objective for the candidate push content based on the initial resource consumption factor of the reference object and the first resource consumption of the reference object for the candidate push content.

[0015] In some embodiments, the candidate resource consumption factor obtaining module is further configured to, for each of the at least two business objectives, obtain a first correlation degree characterization value between the target object and the business objective and a second correlation degree characterization value between the candidate push content and the business objective; and, based on the first correlation degree characterization value and the second correlation degree characterization value, determine the initial resource consumption factor of the candidate push content on the business objective for the target object.

[0016] In some embodiments, the target resource consumption obtaining module is further configured to obtain the actual total resource consumption and the expected total resource consumption of the candidate push content within a second historical time period; adjust the target resource consumption factor based on the actual total resource consumption and the expected total resource consumption to obtain an adjusted target resource consumption factor; and obtain the target resource consumption of the candidate push content for the target object based on the adjusted target resource consumption factor and the expected resource consumption of the candidate push content for the target object.

[0017] In some embodiments, the target resource consumption obtaining module is further configured to determine a second factor adjustment coefficient based on the actual total resource consumption and the expected total resource consumption; and adjust the target resource consumption factor using the second factor adjustment coefficient to obtain the adjusted target resource consumption factor.

[0018] In some embodiments, the target resource consumption obtaining module is further configured to determine a first threshold based on the second factor adjustment coefficient; the first threshold is negatively correlated with the second factor adjustment coefficient; when the target resource consumption factor is less than the first threshold, the target resource consumption factor is increased based on the second factor adjustment coefficient to obtain the adjusted target resource consumption factor.

[0019] On the other hand, this application also provides a computer device. The computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the steps in the above-described content push method.

[0020] On the other hand, this application also provides a computer-readable storage medium. The computer-readable storage medium stores a computer program thereon, which, when executed by a processor, implements the steps in the above-described content push method.

[0021] On the other hand, this application also provides a computer program product. The computer program product includes a computer program that, when executed by a processor, implements the steps in the above-described content push method.

[0022] The aforementioned content push method, apparatus, computer equipment, storage medium, and computer program product, for candidate push content targeting a target object, determine the resource consumption factors corresponding to the candidate push content on at least two business objectives, obtaining each candidate resource consumption factor. Based on each candidate resource consumption factor, determine the target resource consumption factor corresponding to the candidate push content. Based on the target resource consumption factor corresponding to the candidate push content and the expected resource consumption of the candidate push content for the target object, obtain the target resource consumption of the candidate push content for the target object. Content push processing is then performed on the target object based on the target resource consumption. Thus, the target resource consumption is obtained based on the target resource consumption factor determined by each candidate resource consumption factor and the expected resource consumption, rather than processing the expected resource consumption using multiple resource consumption factors, thereby reducing the difference between the target resource consumption and the expected resource consumption, i.e., reducing the impact on resource consumption. Attached Figure Description

[0023] Figure 1 This is a diagram illustrating the application environment of the content push method in some embodiments;

[0024] Figure 2 This is a flowchart illustrating the content push method in some embodiments;

[0025] Figure 3 This is a schematic diagram illustrating the principle of calculating target resource consumption in some embodiments;

[0026] Figure 4 This is a schematic diagram illustrating the content push method in some embodiments;

[0027] Figure 5 This is a schematic diagram illustrating the calculation of eCPM in some embodiments;

[0028] Figure 6 The diagram shows the effect of adjusting the resource consumption factor using the second factor adjustment coefficient in some embodiments.

[0029] Figure 7 This is a flowchart illustrating the content push method in some embodiments;

[0030] Figure 8This is a diagram illustrating application scenarios of the content push method in some embodiments;

[0031] Figure 9 This is a diagram illustrating application scenarios of the content push method in some embodiments;

[0032] Figure 10 This is a framework diagram for determining the target resource consumption factor in some embodiments;

[0033] Figure 11 This is a structural block diagram of the content push device in some embodiments;

[0034] Figure 12 These are internal structural diagrams of the computer device in some embodiments;

[0035] Figure 13 This is a diagram showing the internal structure of a computer device in some embodiments. Detailed Implementation

[0036] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0037] The content push method provided in this application embodiment can be applied to, for example, Figure 1 In the application environment shown, terminal 102 communicates with server 104 via a network. A data storage system can store the data that server 104 needs to process. The data storage system can be integrated onto server 104, or it can be located in the cloud or on another server.

[0038] Specifically, server 104 can determine the resource consumption factors corresponding to the candidate push content for the target object at at least two business objectives, obtain each candidate resource consumption factor, determine the target resource consumption factor corresponding to the candidate push content based on each candidate resource consumption factor, and obtain the target resource consumption amount of the candidate push content for the target object based on the target resource consumption factor and the expected resource consumption amount of the candidate push content for the target object. Then, content push processing is performed on the target object based on the target resource consumption amount. For example, there can be multiple candidate push contents for the target object. Figure 1 When terminal 102 is the target terminal, server 104 can determine the target push content from the candidate push content based on the target resource consumption of each candidate push content, and push the target push content to terminal 102.

[0039] The terminal 102 can be, but is not limited to, various desktop computers, laptops, smartphones, tablets, IoT devices, smart voice interaction devices, smart home appliances, in-vehicle terminals, aircraft, and portable wearable devices. IoT devices can include smart speakers, smart TVs, smart air conditioners, and smart in-vehicle devices. Portable wearable devices can include smartwatches, smart bracelets, and head-mounted devices. The server 104 can be implemented using a standalone server or a server cluster consisting of multiple servers.

[0040] In some embodiments, such as Figure 2 As shown, a content push method is provided. This method can be executed by a terminal or a server, or it can be executed by both a terminal and a server. This method can be applied to... Figure 1 Taking server 104 as an example, the following steps are included:

[0041] Step 202: For the candidate push content of the target object, determine the resource consumption factors corresponding to the candidate push content on at least two business objectives respectively, and obtain each candidate resource consumption factor.

[0042] In this context, "object" represents a user, and the target object can be any object, such as the object to which content is to be pushed. The pushed content can be any content that can be pushed, including at least one of video, audio, image, or text, such as articles, songs, movies, or advertisements (Ads). Candidate push content refers to potential push content. Each object can have one or more candidate push contents, meaning at least two. Different objects can have the same or different candidate push contents. For example, if the push content is an advertisement, advertisement A can be both a candidate push content for object 1 and a candidate push content for object 2. The candidate push contents of the target object are used to determine the target push content to be pushed to the target object. The target push content of the target object belongs to the candidate push contents of the target object. The candidate push contents of the target object can be continuously updated. For example, at a previous moment, the candidate push contents of the target object were candidate push content 1 and candidate push content 2, while at the current moment, the candidate push contents of the target object are candidate push content 1 and candidate push content 3.

[0043] Business objectives are goals set based on specific business objectives. These objectives can be the same or different for different businesses. For example, for a gaming business, the objective might be "to increase the number of paid ads for games." Business objectives can also be called strategies. Resource consumption factors are factors used to adjust resource consumption. There is a one-to-one correspondence between business objectives and resource consumption factors. Candidate resource consumption factors refer to the resource consumption factors corresponding to candidate push content for specific business objectives. Candidate push content can correspond to at least two business objectives. These objectives can be preset, and different candidate push content can have the same or different business objectives. In different scenarios, the resource consumption factors corresponding to the same candidate push content for the same business objective may also be different. The resource consumption factors corresponding to at least two business objectives for candidate push content can refer to the resource consumption factors corresponding to each of the candidate push content's respective business objectives.

[0044] Specifically, for each candidate push content of a target object, the server can determine the resource consumption factor corresponding to that candidate push content on the corresponding business objective, thus obtaining each candidate resource consumption factor. For example, if the target object has two candidate push content, candidate push content 1 and candidate push content 2, and the business objectives corresponding to candidate push content 1 are business objective 1 and business objective 2, and the business objectives corresponding to candidate push content 2 are business objective 3 and business objective 4, then for candidate push content 1, the server can obtain the resource consumption factor 1 of candidate push content 1 on business objective 1 and the resource consumption factor 2 of candidate push content 1 on business objective 2, and determine resource consumption factor 1 and resource consumption factor 2 as each candidate resource consumption factor corresponding to candidate push content 1. Similarly, for candidate push content 2, the server can obtain the resource consumption factor 3 of candidate push content 2 on business objective 3 and the resource consumption factor 4 of candidate push content 2 on business objective 4, and determine resource consumption factor 3 and resource consumption factor 4 as each candidate resource consumption factor of candidate push content 2.

[0045] In some embodiments, taking a candidate push content as an example, the candidate push content for the target object corresponds to an initial resource consumption factor for each corresponding business objective. The server can determine the initial resource consumption factor corresponding to the candidate push content for each corresponding business objective, thus obtaining each candidate resource consumption factor. Alternatively, the server can adjust the initial resource consumption factor to obtain the candidate resource consumption factors. The initial resource consumption factor can be preset or determined based on the degree of correlation between the target object and the business objective, or the degree of correlation between the candidate push content and the business objective.

[0046] Step 204: Determine the target resource consumption factor corresponding to the candidate push content based on each candidate resource consumption factor.

[0047] Specifically, the server can select the target resource consumption factor corresponding to the candidate push content from among the candidate resource consumption factors based on the data distribution information of each candidate resource consumption factor. The data distribution information is used to characterize the data distribution of each candidate resource consumption factor. Specifically, the data distribution information of each candidate resource consumption factor may include at least one of the following: the number of candidate resource consumption factors greater than a preset value, or the number of candidate resource consumption factors less than a preset value. The preset value can be set as needed, for example, to 1. For example, the server can select the target resource consumption factor corresponding to the candidate push content from among the candidate resource consumption factors based on at least one of the following: the number of candidate resource consumption factors greater than a preset value, or the number of candidate resource consumption factors less than a preset value. Of course, the server can also use majority voting, relative majority voting, or weighted voting to select the target resource consumption factor corresponding to the candidate push content from among the candidate resource consumption factors.

[0048] In some embodiments, the server may perform a weighted calculation on each candidate resource consumption factor and determine the result of the weighted calculation as the target resource consumption factor corresponding to the candidate push content. Alternatively, the server may determine the average value of each candidate resource consumption factor as the target resource consumption factor corresponding to the candidate push content.

[0049] Step 206: Based on the target resource consumption factor corresponding to the candidate push content and the expected resource consumption of the candidate push content for the target object, obtain the target resource consumption of the candidate push content for the target object; the expected resource consumption is the expected resource consumption when the push object of the candidate push content is the target object.

[0050] Resource consumption refers to the resources required to push content to the target audience, such as the resources required to run an advertisement. Expected resource consumption is the anticipated resource consumption when the target audience for the candidate push content is the intended audience. The target audience refers to the individuals to whom the content is pushed. The calculation method for expected resource consumption can be set as needed. For example, taking an advertisement as an example, the expected resource consumption can be eCPM (effective cost per mille). eCPM refers to the amount of resources a traffic owner can obtain per thousand impressions (exposures), or it can represent the amount of resources an advertiser needs to expend per thousand impressions (exposures). eCPM = Estimated Click-Through Rate * Estimated Conversion Rate * Target Conversion Bid * 1000. The estimated click-through rate is the estimated probability that the target audience will click on the advertisement, and the estimated conversion rate is the estimated probability that the target audience will convert after clicking on the advertisement. Conversion can be determined based on business objectives. For example, if the business objective is to increase purchase volume, then if the target audience makes a purchase, it can be determined that the target audience has converted.

[0051] Specifically, the server can obtain the target resource consumption of the candidate push content for the target object based on the target resource consumption factor corresponding to the candidate push content and the expected resource consumption of the candidate push content for the target object. The target resource consumption is positively correlated with the expected resource consumption, and the target resource consumption is positively correlated with the target resource consumption factor.

[0052] In some embodiments, the server may multiply the target resource consumption factor by the expected resource consumption and determine the result of the operation as the target resource consumption of the candidate push content for the target object.

[0053] In some embodiments, the server may adjust the target resource consumption factor to obtain the adjusted target resource consumption factor, and based on the adjusted target resource consumption factor and the expected resource consumption of the candidate push content for the target object, obtain the target resource consumption of the candidate push content for the target object.

[0054] Step 208: Perform content push processing on the target object based on the target resource consumption.

[0055] Specifically, there can be multiple candidate push content items for a target object, meaning at least two. The server can obtain the target resource consumption corresponding to each candidate push content item for the target object. Based on the target resource consumption of each candidate push content item, the server selects the target push content from the candidate push content items and pushes the target push content to the target object's terminal, where the target object's terminal can display the target push content. Resource consumption reflects the cost required to push content.

[0056] In some embodiments, the server can select the candidate push content corresponding to the largest target resource consumption from among the candidate push content of the target object to obtain the target push content. Alternatively, the server can determine the candidate push content whose target resource consumption is greater than a resource consumption threshold as the target push content. The resource consumption threshold can be preset or set as needed.

[0057] In some embodiments, the server can arrange the candidate push content in descending order of target resource consumption to obtain a candidate push content sequence. The greater the target resource consumption, the higher the ranking of the candidate push content in the candidate push content sequence. The server can determine the candidate push content whose content is ranked before a ranking threshold as the target push content. Content ranking refers to the order of candidate push content in the candidate push content sequence. The ranking threshold can be preset or set as needed, for example, it can be any one of 2 or 4.

[0058] In some embodiments, the business objective can be referred to as a strategy, and the resource consumption factor can be referred to as a personalized bidding factor. Taking the content push method of this application as an example, with a resource consumption of eCPM, three candidate push contents for the target audience (Ad1, Ad2, and Ad3), and the strategies corresponding to these three candidate push contents being Strategy 1, Strategy 2, and Strategy 3 respectively, as an example... Figure 3 As shown, "Ad1 eCPM=30" indicates that the expected resource consumption of candidate push content Ad1 is 30, "Ad2 eCPM=20" indicates that the expected resource consumption of candidate push content Ad2 is 20, and "Ad3 eCPM=10" indicates that the expected resource consumption of candidate push content Ad3 is 10. For strategy 1, the candidate resource consumption factor for Ad1 is 1.5, for Ad2 it is 0.8, and for Ad3 it is 1.5. For strategy 2, the candidate resource consumption factor for Ad1 is 1.2, for Ad2 it is 0.6, and for Ad3 it is 0.55. For strategy 3, the candidate resource consumption factor for Ad1 is 1.1, for Ad2 it is 0.5, and for Ad3 it is 1.2. Figure 3The unified algorithm mechanism refers to the method for determining the target resource consumption factor. Using this mechanism, the target resource consumption factor for Ad1 is determined to be 1.5, for Ad2 it is 0.5, and for Ad3 it is 1.5. After obtaining the target resource consumption factors for these three candidate push contents, the target resource consumption factor (i.e., 1.5) of candidate push content Ad1 is multiplied by the expected resource consumption of Ad1 (i.e., 30) to obtain the target resource consumption of Ad1 (i.e., 45). The target resource consumption factor (0.5) of candidate push content Ad2 is multiplied by the expected resource consumption of Ad2 (20) to obtain the target resource consumption of Ad2 (10). The target resource consumption factor (1.5) of candidate push content Ad3 is multiplied by the expected resource consumption of Ad3 (10) to obtain the target resource consumption of Ad3 (15). Figure 3 In the text, “Ad1 eCPM=45” means that the target resource consumption of candidate push content Ad1 is 45, “Ad2 eCPM=10” means that the expected resource consumption of candidate push content Ad2 is 10, and “Ad3 eCPM=15” means that the expected resource consumption of candidate push content Ad3 is 15.

[0059] In the aforementioned content push method, for candidate push content targeting a target audience, resource consumption factors corresponding to each candidate push content are determined for at least two business objectives, resulting in each candidate resource consumption factor. Based on each candidate resource consumption factor, a target resource consumption factor corresponding to the candidate push content is determined. Based on the target resource consumption factor corresponding to the candidate push content and the expected resource consumption of the candidate push content for the target audience, the target resource consumption of the candidate push content for the target audience is obtained. Content push processing is then performed on the target audience based on the target resource consumption factor determined by each candidate resource consumption factor and the expected resource consumption, rather than processing the expected resource consumption using multiple resource consumption factors. This reduces the difference between the target resource consumption and the expected resource consumption, thus minimizing the impact on resource consumption.

[0060] In some embodiments, determining the target resource consumption factor corresponding to the candidate push content based on each candidate resource consumption factor includes: determining the data distribution information of each candidate resource consumption factor; and selecting the target resource consumption factor corresponding to the candidate push content from each candidate resource consumption factor based on the data distribution information.

[0061] The data distribution information for each candidate resource consumption factor may include at least one of the following: the number of candidate resource consumption factors that are greater than a preset value, or the number of candidate resource consumption factors that are less than the preset value. The preset value can be set as needed, for example, to 1.

[0062] Specifically, the server can count the number of candidate resource consumption factors that are greater than a preset value and the number of candidate resource consumption factors that are less than the preset value, thereby obtaining the data distribution information of each candidate resource consumption factor.

[0063] In this embodiment, based on data distribution information, the target resource consumption factor corresponding to the candidate push content is selected from each candidate resource consumption factor. Thus, based on the statistically obtained data, the target resource consumption factor corresponding to the candidate push content is selected from each candidate resource consumption factor, which improves the rationality and accuracy of screening target resource consumption factors.

[0064] In some embodiments, selecting the target resource consumption factor corresponding to the candidate push content from each candidate resource consumption factor based on data distribution information includes: determining the number of candidate resource consumption factors greater than a preset value among each candidate resource consumption factor to obtain a first number; determining the number of candidate resource consumption factors less than a preset value among each candidate resource consumption factor to obtain a second number; and selecting the target resource consumption factor corresponding to the candidate push content from each candidate resource consumption factor based on the first number and the second number.

[0065] The first quantity refers to the number of candidate resource consumption factors that are greater than a preset value. The second quantity refers to the number of candidate resource consumption factors that are less than a preset value. The preset value can be set as needed, for example, to 1. The data distribution information includes the first and second quantities.

[0066] Specifically, the server can count the number of candidate resource consumption factors that are greater than a preset value among the candidate resource consumption factors to obtain a first number, count the number of candidate resource consumption factors that are less than a preset value among the candidate resource consumption factors to obtain a second number, and select the target resource consumption factor corresponding to the candidate push content from among the candidate resource consumption factors based on the first number and the second number.

[0067] In some embodiments, the server may compare the first quantity with the second quantity, and based on the comparison result, select the target resource consumption factor corresponding to the candidate push content from each candidate resource consumption factor.

[0068] In this embodiment, since both the first quantity and the second quantity are statistical data, the target resource consumption factor corresponding to the candidate push content is selected from each candidate resource consumption factor based on the first quantity and the second quantity. This improves the rationality and accuracy of the selection of target resource consumption factors.

[0069] In some embodiments, selecting the target resource consumption factor corresponding to the candidate push content from each candidate resource consumption factor based on the first quantity and the second quantity includes: determining a factor screening strategy based on the first quantity and the second quantity when the first quantity and the second quantity are different; and selecting the target resource consumption factor corresponding to the candidate push content from each candidate resource consumption factor based on the factor screening strategy.

[0070] The factor selection strategy may include at least one of a maximum factor selection strategy or a minimum factor selection strategy. The maximum factor selection strategy is the strategy that selects the candidate resource consumption factor with the largest value, while the minimum factor selection strategy is the strategy that selects the candidate resource consumption factor with the smallest value.

[0071] Specifically, the server can compare the first quantity with the second quantity. If the first quantity is greater than the second quantity, the factor selection strategy is determined to be the maximum factor selection strategy, selecting the largest candidate resource consumption factor from all candidate resource consumption factors, and determining the selected candidate resource consumption factor as the target resource consumption factor for the candidate push content. If the first quantity is less than the second quantity, the factor selection strategy is determined to be the minimum factor selection strategy, selecting the smallest candidate resource consumption factor from all candidate resource consumption factors, and determining the selected candidate resource consumption factor as the target resource consumption factor for the candidate push content.

[0072] In some embodiments, the server can compare the first quantity with the second quantity. If the first quantity and the second quantity are the same, the target resource consumption factor is determined to be a preset resource consumption factor. The preset resource consumption factor can be preset as needed, for example, the preset resource consumption factor can be 1.

[0073] In some embodiments, candidate resource consumption factors greater than a preset value can be called upward adjustment factors, and candidate resource consumption factors less than a preset value can be called downward adjustment factors. Therefore, the first quantity can be called the quantity of upward adjustment factors, and the second quantity can be called the quantity of downward adjustment factors. Taking a preset value of 1 as an example, the quantity m of upward adjustment factors can be expressed as formula (1), and the quantity n of downward adjustment factors can be expressed as formula (2).

[0074]

[0075] Where m represents the number of upward adjustment factors (the first number), n represents the number of downward adjustment factors (the second number), 1 represents the preset value, and score′ i Let represent the i-th candidate resource consumption factor, and T represent the number of candidate resource consumption factors for the target object, for example, 3. I takes the value 1 or 0. In formula (1), when score′ iWhen I is greater than 1, I takes the value of 1; when score′ i When I is less than 1, I takes the value of 0. In formula (2), when score′ i When I is greater than 1, I takes the value of 0; when score′ i When I is less than 1, I takes the value 1.

[0076] In some embodiments, when the first quantity and the second quantity are different, the server can determine the adjustment direction based on the comparison result, determine the factor screening strategy based on the adjustment direction, and select the target resource consumption factor corresponding to the candidate push content from each candidate resource consumption factor based on the factor screening strategy. The adjustment direction can be any one of upward adjustment, downward adjustment, or no adjustment.

[0077] In some embodiments, when the first quantity is greater than the second quantity, the server determines the adjustment direction to be upward. If the adjustment direction is upward, the factor selection strategy is determined to be the maximum factor selection strategy, and the largest candidate resource consumption factor among all candidate resource consumption factors is determined as the target resource consumption factor for the candidate push content. When the first quantity is less than the second quantity, the server determines the adjustment direction to be downward. If the adjustment direction is downward, the factor selection strategy is determined to be the minimum factor selection strategy, and the smallest candidate resource consumption factor among all candidate resource consumption factors is determined as the target resource consumption factor for the candidate push content. When the first quantity and the second quantity are the same, the server determines the adjustment direction to be no adjustment, and a preset resource consumption factor is used as the target resource consumption factor. Figure 3 As shown, the three candidate resource consumption factors for candidate push content Ad1 are 1.5, 1.2, and 1.1. Since there are three candidate resource consumption factors greater than 1 and zero candidate resource consumption factors less than 1, the adjustment direction is upward. The largest value among 1.5, 1.2, and 1.1, 1.5, is determined as the target resource consumption factor for candidate push content Ad1. Figure 3 In the text, an upward arrow represents an increase, and a downward arrow represents a decrease.

[0078] The method for determining the target resource consumption factor provided in this embodiment has the advantage of convenient attribution. For example, when the adjustment direction is determined to be upward, it can first be attributed to the factor with the largest value, and secondly to the remaining upward adjustment factors (the remaining upward adjustment factors at least contributed to the direction). Moreover, the withdrawal of a factor does not affect other factors. For example, if a factor is set to a smaller upward adjustment range due to poor performance, it will not affect the result if other factors are larger. The factor refers to the candidate resource consumption factor. For example, taking the preset resource consumption factor as 1 as an example, the adjustment direction H can be represented by formula (3), and the target resource consumption factor score″ can be represented by formula (4). i .

[0079]

[0080] Where H represents the adjustment direction, H=1 indicates an upward adjustment, H=0 indicates a downward adjustment, and H=-1 indicates no adjustment. i Represents the target resource consumption factor. max{score′ i ,with score′ i >1} represents the largest candidate resource consumption factor among all candidate resource consumption factors, min{score′ i ,with score′ i <1} represents the smallest candidate resource consumption factor among all candidate resource consumption factors.

[0081] In this embodiment, when the first quantity and the second quantity are different, a factor screening strategy is determined based on the first quantity and the second quantity. Based on the factor screening strategy, the target resource consumption factor corresponding to the candidate push content is selected from each candidate resource consumption factor. Thus, the target resource consumption factor can be determined according to the size relationship between the first quantity and the second quantity, which improves the rationality and accuracy of determining the target resource consumption factor.

[0082] In some embodiments, determining the resource consumption factors corresponding to candidate push content on at least two business objectives to obtain each candidate resource consumption factor includes: determining the initial resource consumption factors corresponding to candidate push content on at least two business objectives for a target object; for each of the at least two business objectives, obtaining a first factor adjustment coefficient for the candidate push content; and adjusting the initial resource consumption factors of the candidate push content on the business objective using the first factor adjustment coefficient for the candidate push content to obtain the candidate resource consumption factors of the candidate push content on the business objective.

[0083] The initial resource consumption factor can be preset or determined based on at least one of the correlation between the target object and the business objective, or the correlation between the candidate push content and the business objective. The correlation between the target object and the business objective is positively correlated with the initial resource consumption factor, and the correlation between the candidate push content and the business objective is also positively correlated with the initial resource consumption factor. The first factor adjustment coefficient for the business objective regarding the candidate push content refers to the first factor adjustment coefficient jointly determined by the business objective and the candidate push content. Each business objective and each candidate push content uniquely determines one first factor adjustment coefficient. For example, business objective 1 and candidate push content Ad1 uniquely determine one first factor adjustment coefficient, which is the first factor adjustment coefficient of business objective 1 for candidate push content Ad1.

[0084] Specifically, for candidate push content for a target object, the server can determine the initial resource consumption factor corresponding to the candidate push content for each relevant business objective. For each business objective, the initial resource consumption factor of the candidate push content is adjusted using the first factor adjustment coefficient for the business objective, resulting in the candidate resource consumption factor of the candidate push content for the business objective. The server can calculate the ratio of the initial resource consumption factor to the first factor adjustment coefficient and determine this ratio as the candidate resource consumption factor of the candidate push content for the business objective. For example, the formula score′ can be used. i =score i / x(5) calculates the candidate resource consumption factor. In formula (5), score i represents the initial resource consumption factor, and x represents the adjustment coefficient of the first factor.

[0085] In some embodiments, the first factor adjustment coefficient can be continuously updated. For example, the initial first factor adjustment coefficient can be a preset factor adjustment coefficient, which can be pre-set as needed, for example, the preset factor adjustment coefficient can be 1. The server can continuously update the first factor adjustment coefficient and use the updated first factor adjustment coefficient to adjust the initial resource consumption factor of the candidate push content on the business objective, thereby obtaining the candidate resource consumption factor of the candidate push content on the business objective, and continuously looping. For example, the server can update the first factor adjustment coefficient every preset time interval. The preset time interval can be pre-set as needed, for example, the preset time interval can be 2 minutes. Figure 4 As shown, there are four strategies for the candidate push content: Strategy 1, Strategy 2, Strategy 3, and Strategy 4. Figure 4The personalized bidding constraint refers to the method for determining and updating the first factor adjustment coefficient. "Strategy1_score" is the initial resource consumption factor of the candidate content under Strategy 1, "Strategy2_score" is the initial resource consumption factor of the candidate content under Strategy 2, "Strategy3_score" is the initial resource consumption factor of the candidate content under Strategy 3, and "Strategy4_score" is the initial resource consumption factor of the candidate content under Strategy 4. "Strategy1_score'" is the candidate resource consumption factor obtained by adjusting "Strategy1_score" using the first factor adjustment coefficient, "Strategy2_score'" is the candidate resource consumption factor obtained by adjusting "Strategy2_score" using the first factor adjustment coefficient, "Strategy3_score'" is the candidate resource consumption factor obtained by adjusting "Strategy3_score" using the first factor adjustment coefficient, and "Strategy4_score'" is the candidate resource consumption factor obtained by adjusting "Strategy4_score" using the first factor adjustment coefficient. Each strategy corresponds one-to-one with the first factor adjustment coefficient. Figure 4 From the four candidate resource consumption factors, "Strategy3_score" was selected as the target resource consumption factor.

[0086] In this embodiment, the initial resource consumption factor of the candidate push content is adjusted based on the first factor adjustment coefficient of the business objective, thereby obtaining the candidate resource consumption factor of the candidate push content on the business objective. This allows the resource consumption factor to be influenced according to the business objective, making the resource consumption factor more in line with the needs of the business objective.

[0087] In some embodiments, there are multiple target objects, and each target object corresponds to multiple candidate push content. The method further includes: for each target object, selecting target push content corresponding to the target object from the multiple candidate push content corresponding to the target object based on the target resource consumption of the target object; for each target object, pushing the target push content corresponding to the target object to the target object's terminal; updating the first factor adjustment coefficients of at least two business objectives for the candidate push content based on multiple historical push records within a first historical time period; and returning to the step of adjusting the initial resource consumption factor of the candidate push content on the business objective using the first factor adjustment coefficients of the business objective for the candidate push content to obtain the candidate resource consumption factor of the candidate push content on the business objective.

[0088] In this context, there are multiple target objects, with "each" referring to at least two. For each target object and each candidate push content, the server can determine the target resource consumption of the candidate push content for that target object using the methods in steps 202-208. The first historical time period can be determined based on the update cycle of the first factor adjustment coefficient. For example, if the first factor adjustment coefficient is updated every 2 minutes, then the first historical time period can be the most recent 2 minutes. Historical push records are used to record information about a request within a first historical time period. A request represents the interaction between the target object and the corresponding target push content, and the interaction includes, but is not limited to, at least one of exposure, click, or conversion. The historical push record corresponding to a request can include the identifier of the business target, the identifier of the push content, the identifier of the object, the interaction behavior, the interaction time, etc. The identifier of the business objective recorded in the historical push record refers to the identifier of the target resource consumption factor of the push content represented by the identifier of the push content recorded in the historical push record. For example, if the historical push record includes the identifier of business objective 1, the identifier of push content 1, and the identifier of object 1, then it can be determined that: before pushing push content 1 to the terminal of object 1, the target resource consumption factor corresponding to push content 1 determined by the server from each candidate resource consumption factor is the resource consumption factor of push content 1 on business objective 1.

[0089] Specifically, once the server has determined the target content for the target audience, it can push that content to the target audience's device. After a period of time, the server can update the first factor adjustment coefficient based on historical push data. For example, the server can update the first factor adjustment coefficient periodically, such as every 2 minutes.

[0090] In some embodiments, the server may update the first factor adjustment coefficient for each target service for each candidate push content based on multiple historical push records within a first historical time period. After adjusting the first factor adjustment coefficient for each target service for each candidate push content, the server may return to the step of adjusting the initial resource consumption factor of the candidate push content on the business target using the first factor adjustment coefficient for the candidate push content of the business target, thereby obtaining the candidate resource consumption factor of the candidate push content on the business target.

[0091] In this embodiment, for each target object, target push content corresponding to the target object is pushed to the target object's terminal. Based on multiple historical push records within a first historical time period, the first factor adjustment coefficients for at least two business objectives regarding the candidate push content are updated. This allows for continuous correction of the value of the first factor adjustment coefficients to promptly respond to updates to the candidate push content of the target object, thereby improving the adaptability of the first factor adjustment coefficients. For example, in an advertising bidding scenario, the candidate push content can be called bidding advertisements. As time changes, the bidding advertisements of the target object may also change continuously. Timely updating the first factor adjustment coefficients can promptly apply changes in the bidding environment, which may include changes in the various bidding advertisements fulfilled by the target, thus enabling the first factor adjustment coefficients to adapt to changes in the bidding environment.

[0092] In some embodiments, updating the first factor adjustment coefficients for candidate push content for at least two business objectives based on multiple historical push records within a first historical time period includes: for each of the at least two business objectives, selecting historical push records containing the identifier of the candidate push content and the identifier of the business objective from the multiple historical push records to obtain a reference push record corresponding to the business objective; determining a reference object corresponding to the business objective based on the object identifier recorded in each reference push record corresponding to the business objective; determining the initial resource consumption factor of the candidate push content on the business objective for the reference object; and updating the first factor adjustment coefficients for the candidate push content for the business objective based on the initial resource consumption factor.

[0093] Here, the reference object refers to the object represented by the object identifier recorded in the reference push record. Specifically, taking candidate push content 1 and business objective 1 as an example, for each reference object corresponding to business objective 1, the server can determine the initial resource consumption factor of candidate push content 1 on business objective 1, obtain the first resource consumption amount of the reference object for candidate push content 1, and update the first factor adjustment coefficient of business objective 1 for candidate push content 1 based on the initial resource consumption factor and the first resource consumption amount. The first resource consumption amount of the reference object for candidate push content 1 refers to the expected resource consumption amount for pushing candidate push content 1 to the reference object under a preset resource consumption strategy. The preset resource consumption strategy can be either calculating resource consumption based on exposure or calculating resource consumption based on clicks. The calculation method for the first resource consumption amount can be the same as or different from the calculation method for the expected resource consumption amount.

[0094] Specifically, for each reference push record, the server can adjust the first resource consumption based on the initial resource consumption factor of the reference object corresponding to that reference push record, thus obtaining the adjusted first resource consumption for that reference push record. For example, the initial resource consumption factor can be multiplied by the first resource consumption, and the result of the multiplication can be determined as the adjusted first resource consumption. The server can update the first factor adjustment coefficient of the business objective for the candidate push content based on the adjusted first resource consumption corresponding to each reference push record.

[0095] In some embodiments, the server can sum the adjusted first resource consumption corresponding to each reference push record to obtain a first total resource consumption. Based on the first total resource consumption, the server updates the first factor adjustment coefficient of the business objective for the candidate push content. The updated first factor adjustment coefficient is positively correlated with the first total resource consumption.

[0096] In this embodiment, the initial resource consumption factor of the candidate push content on the business objective is determined for the reference object. Based on the initial resource consumption factor, the first factor adjustment coefficient of the business objective for the candidate push content is updated, so that the first adjustment factor coefficient can change according to the changes in the environment, thereby improving the applicability and accuracy of the first factor adjustment coefficient.

[0097] In some embodiments, updating the first factor adjustment coefficient of the business objective for the candidate push content based on the initial resource consumption factor includes: obtaining the first resource consumption of the reference object corresponding to the business objective for the candidate push content; the first resource consumption refers to the expected resource consumption under the preset resource consumption strategy; and updating the first factor adjustment coefficient of the business objective for the candidate push content based on the initial resource consumption factor of the reference object and the first resource consumption of the reference object for the candidate push content.

[0098] Specifically, the server can sum the adjusted first resource consumption corresponding to each reference push record to obtain the first total resource consumption. Here, the adjusted first resource consumption represents the resource consumption generated under the preset resource consumption strategy after adjusting the initial resource consumption factor. The server can sum the first resource consumption of the reference objects corresponding to each reference push record to obtain the second total resource consumption. Based on the first and second total resource consumption, the server can calculate a new first factor adjustment coefficient. The new first factor adjustment coefficient is positively correlated with the first total resource consumption and negatively correlated with the second total resource consumption. For example, the new first factor adjustment coefficient can be expressed by formula (6).

[0099]

[0100] Where x represents the new first factor adjustment coefficient, bid_price j The score represents the first resource consumption corresponding to the reference object of the j-th reference push record representing the business objective. j The initial resource consumption factor represents the reference object corresponding to the j-th reference push record corresponding to the business objective. Since the first resource consumption refers to the expected resource consumption under the preset resource consumption strategy, and the adjusted first resource consumption represents the resource consumption generated under the preset resource consumption strategy after the adjustment of the initial resource consumption factor, the new first factor adjustment coefficient can be obtained by using formula (6), so that the difference between the adjusted resource consumption and the expected resource consumption can be minimized. This can be seen from formula (7).

[0101]

[0102] Where, ∑ j bid_price j *score j / x represents the resource consumption under the influence of x, ∑ j bid_price j This represents the amount of resources consumed without the influence of x (i.e., the expected amount of resources consumed). It can be seen that the amount of resources consumed is the same whether or not x is present.

[0103] In this embodiment, based on the initial resource consumption factor of the reference object and the first resource consumption of the reference object for the candidate push content, the first factor adjustment coefficient of the business target for the candidate push content is updated. This can minimize the change in resource consumption caused by the first factor adjustment coefficient, thereby reducing the impact on resource consumption.

[0104] In some embodiments, determining the initial resource consumption factor for the candidate push content on at least two business objectives for the target object includes: for each of the at least two business objectives, obtaining a first correlation degree representation value between the target object and the business objective and a second correlation degree representation value between the candidate push content and the business objective; and determining the initial resource consumption factor for the candidate push content on the business objective for the target object based on the first correlation degree representation value and the second correlation degree representation value.

[0105] The first correlation degree characterization value represents the degree of correlation between the target object and the business objective. A larger first correlation degree characterization value indicates a stronger correlation between the target object and the business objective. The second correlation degree characterization value represents the degree of correlation between the candidate push content and the business objective. A larger second correlation degree characterization value indicates a stronger correlation between the candidate push content and the business objective. For example, if the business objective is "to increase the paid subscription rate for game ads," target 1 is an object that likes playing games, the candidate push content is game ads, and target 2 is an object that does not like playing games, then the correlation between target 1 and the business objective is greater than the correlation between target 2 and the business objective. The initial resource consumption factor is positively correlated with the first correlation degree characterization value, and the initial resource consumption factor is positively correlated with the second correlation degree characterization value.

[0106] Specifically, the server can multiply the first association degree representation value and the second association degree representation value, and determine the result as the initial resource consumption factor of the candidate push content for the target object in terms of business objectives.

[0107] In this embodiment, based on the first correlation degree characterization value and the second correlation degree characterization value, the initial resource consumption factor of the candidate push content on the business objective is determined for the target object. Thus, the initial resource consumption factor is determined according to the correlation degree between the business objective, the target object, and the candidate push content, thereby improving the rationality and accuracy of the initial resource consumption factor.

[0108] In some embodiments, obtaining the target resource consumption of the candidate push content for the target object based on the target resource consumption factor corresponding to the candidate push content and the expected resource consumption of the candidate push content for the target object includes: obtaining the actual total resource consumption and the expected total resource consumption of the candidate push content in a second historical time period; adjusting the target resource consumption factor based on the actual total resource consumption and the expected total resource consumption to obtain the adjusted target resource consumption factor; and obtaining the target resource consumption of the candidate push content for the target object based on the adjusted target resource consumption factor and the expected resource consumption of the candidate push content for the target object.

[0109] The second historical time period can be set as needed, for example, it can be the most recent 10 minutes. The second historical time period can be the same as or different from other historical time periods. The actual total resource consumption of the candidate push content within the second historical time period refers to the total resources actually consumed by pushing the candidate push content within the second historical time period. The expected total resource consumption of the candidate push content within the second historical time period refers to the total resources expected to be consumed by pushing the candidate push content within the second historical time period.

[0110] Specifically, for each candidate push content, the server can periodically determine the actual total resource consumption and the expected total resource consumption of the candidate push content within a second historical time period. For example, every 10 minutes, the server can determine the actual total resource consumption and the expected total resource consumption of the candidate push content within the most recent 10 minutes, calculate the difference between the actual total resource consumption and the expected total resource consumption, and define this difference as the consumption difference. If the consumption difference is greater than a preset difference threshold, the target resource consumption factor is adjusted based on the actual total resource consumption and the expected total resource consumption to obtain the adjusted target resource consumption factor. Based on the adjusted target resource consumption factor and the expected resource consumption of the candidate push content for the target object, the target resource consumption of the candidate push content for the target object is obtained. The preset difference threshold can be predetermined as needed and is a value greater than 0.

[0111] In some embodiments, the server can determine a second factor adjustment coefficient based on the actual total resource consumption and the expected total resource consumption. The second factor adjustment coefficient is positively correlated with the actual total resource consumption and negatively correlated with the expected total resource consumption. The server can adjust the target resource consumption factor based on the second factor adjustment coefficient to obtain the adjusted target resource consumption factor.

[0112] In some embodiments, the server may multiply the adjusted target resource consumption factor with the expected resource consumption of the candidate push content for the target object, and determine the result of the calculation as the target resource consumption of the candidate push content for the target object.

[0113] In this embodiment, the target resource consumption factor is adjusted based on the actual total resource consumption and the expected total resource consumption to obtain the adjusted target resource consumption factor. Based on the adjusted target resource consumption factor and the expected resource consumption of the candidate push content for the target object, the target resource consumption of the candidate push content for the target object is obtained. Thus, if the resource consumption is significantly affected by the adjustment of the first factor adjustment coefficient, the resource consumption can be further adjusted to reduce the impact on resource consumption.

[0114] In some embodiments, adjusting the target resource consumption factor based on the actual total resource consumption and the expected total resource consumption to obtain the adjusted target resource consumption factor includes: determining a second factor adjustment coefficient based on the actual total resource consumption and the expected total resource consumption; and adjusting the target resource consumption factor using the second factor adjustment coefficient to obtain the adjusted target resource consumption factor.

[0115] Specifically, the server can calculate the ratio of actual total resource consumption to expected total resource consumption, determine the calculated ratio as the second factor adjustment coefficient, and use this second factor adjustment coefficient to adjust the target resource consumption factor to obtain the adjusted target resource consumption factor. For example, the server can multiply the second factor adjustment coefficient and the target resource consumption factor, and determine the result as the adjusted target resource consumption factor. Alternatively, the server can calculate the ratio of the target resource consumption factor to the second factor adjustment coefficient, and determine the calculated ratio as the adjusted target resource consumption factor.

[0116] In some embodiments, when the difference in consumption is greater than a preset difference threshold, the server can use the ratio of the actual total resource consumption to the expected total resource consumption to determine the calculated ratio as the second factor adjustment coefficient, and use the second factor adjustment coefficient to adjust the target resource consumption factor to obtain the adjusted target resource consumption factor.

[0117] In this embodiment, a second factor adjustment coefficient is determined based on the actual total resource consumption and the expected total resource consumption. The target resource consumption factor is then adjusted using the second factor adjustment coefficient to obtain the adjusted target resource consumption factor. Thus, if the resource consumption is significantly affected by the adjustment of the first factor adjustment coefficient, the resource consumption can be further adjusted using the second factor adjustment coefficient to reduce the impact on resource consumption.

[0118] In some embodiments, adjusting the target resource consumption factor using the second factor adjustment coefficient to obtain the adjusted target resource consumption factor includes: determining a first threshold based on the second factor adjustment coefficient; the first threshold being negatively correlated with the second factor adjustment coefficient; and when the target resource consumption factor is less than the first threshold, increasing the target resource consumption factor based on the second factor adjustment coefficient to obtain the adjusted target resource consumption factor.

[0119] The first threshold and the second factor adjustment coefficient are negatively correlated; for example, the first threshold is the reciprocal of the second factor adjustment coefficient. The second factor adjustment coefficient is the ratio of the actual total resource consumption to the expected total resource consumption. When the consumption difference exceeds a preset difference threshold (which is greater than 0), the actual total resource consumption exceeds the expected total resource consumption, therefore the second factor adjustment coefficient is greater than 1. Since the first threshold is the reciprocal of the second factor adjustment coefficient, it is less than 1.

[0120] Specifically, when the target resource consumption factor is less than the first threshold, the server increases the target resource consumption factor based on the second factor adjustment coefficient to obtain the adjusted target resource consumption factor. For example, the server can multiply the second factor adjustment coefficient with the target resource consumption factor and determine the result as the adjusted target resource consumption factor. This achieves dynamic adjustment of the target resource consumption factor. Figure 4 After obtaining the target resource consumption factor "strategy3_score'", the "strategy3_score'" is dynamically adjusted.

[0121] In some embodiments, when the target resource consumption factor is greater than or equal to a first threshold and less than or equal to a second factor adjustment coefficient, the server may determine the first value as the adjusted target resource consumption factor. The first value can be set as needed, for example, the first value can be 1.

[0122] In some embodiments, when the target resource consumption factor is greater than the second factor adjustment coefficient, the server adjusts the target resource consumption factor by reducing it based on the second factor adjustment coefficient to obtain the adjusted target resource consumption factor. For example, the server can calculate the ratio between the target resource consumption factor and the second factor adjustment coefficient and determine the calculated ratio as the adjusted target resource consumption factor.

[0123] For example, the adjusted target resource consumption factor can be expressed by formula (8).

[0124] f(score”,y)∈[TH low ,TH high (9)

[0125] Where INF is an abbreviation for infimum, representing the infimum. "score" represents the target resource consumption factor, f(score",y) represents the adjusted target resource consumption factor, y represents the second factor adjustment coefficient, and TH... low Let TH be the minimum value of f(score”,y). high Let f(score”,y) be the maximum value.

[0126] In some embodiments, the server can multiply the adjusted target resource consumption factor with the expected resource consumption of the candidate push content for the target object, and determine the result as the target resource consumption of the candidate push content for the target object. For example, taking the expected resource consumption as eCPM, the target resource consumption can be expressed as eCPM' = targetCPA * pCVR * pCTR * f(score”, y) (10). Here, eCPM' is the target resource consumption, targetCPA represents the target conversion bid, pCVR represents the estimated conversion rate, and pCTR represents the estimated click-through rate. Figure 5 As shown, pCTCVR represents the result of multiplying pCVR and pCTR, and F represents the method for selecting personalized bidding factors. The personalized bidding factors corresponding to multiple strategies are input into F in the personalized bidding framework. The output of F is the personalized bidding factor selected from each personalized bidding factor, which is the target resource consumption factor. The eCPM (referring to the target resource consumption) is calculated using this personalized bidding factor.

[0127] In this embodiment, when the target resource consumption factor is less than the first threshold, the target resource consumption factor is increased based on the second factor adjustment coefficient to obtain the adjusted target resource consumption factor. This increases the target resource consumption factor towards 1, thereby reducing the difference between the target resource consumption factor and 1, and thus reducing the impact on resource consumption. Figure 6 The figure illustrates the relationship between the target resource consumption factor before and after adjustment under different second factor adjustment coefficients. As can be seen from the figure, the larger the second factor adjustment coefficient, the closer the adjusted target resource consumption factor is to 1. When the adjusted target resource consumption factor equals 1, it is equivalent to no adjustment being made to the expected resource consumption, thus the obtained target resource consumption is consistent with the expected resource consumption. This embodiment implements a method for dynamically adjusting the target resource consumption factor. Although personalized bidding constraints (i.e., the first factor adjustment coefficient) can make the overall mean of a single factor 1, large upward or downward adjustments can still have a significant impact on GMV (Gross Merchandise Volume). Taking advertising as an example, if advertising costs are severely exceeded, the impact of personalized bidding on GMV may exacerbate the cost overrun. GMV refers to the total transaction volume over a period of time. For example, GMV = target conversion bid * conversion volume, where the target conversion bid is the average conversion cost desired by the advertiser.

[0128] In some embodiments, such as Figure 7As shown, a content push method is provided. This method can be executed by a terminal or a server, or by both a terminal and a server. Taking the application of this method to a server as an example, it includes the following steps:

[0129] Step 702: Determine the initial resource consumption factor for each business objective of the candidate push content for the target object.

[0130] Step 704: For each business objective corresponding to the candidate push content, obtain the first factor adjustment coefficient of the business objective for the candidate push content.

[0131] Step 706: Using the first factor adjustment coefficient of the business objective for the candidate push content, adjust the initial resource consumption factor of the candidate push content on the business objective to obtain the candidate resource consumption factor of the candidate push content on the business objective.

[0132] Step 708: Determine the data distribution information of each candidate resource consumption factor, and select the target resource consumption factor corresponding to the candidate push content from each candidate resource consumption factor based on the data distribution information.

[0133] Step 710: Obtain the actual total resource consumption and the expected total resource consumption of the candidate push content in the second historical time period, and calculate the consumption difference obtained by subtracting the expected total resource consumption from the actual total resource consumption.

[0134] Step 712: Determine whether the consumption difference is greater than the preset difference threshold. If yes, proceed to step 714; otherwise, proceed to step 718.

[0135] Step 714: Adjust the target resource consumption factor based on the actual total resource consumption and the expected total resource consumption to obtain the adjusted target resource consumption factor.

[0136] Step 716: Based on the adjusted target resource consumption factor and the expected resource consumption of the candidate push content for the target object, obtain the target resource consumption of the candidate push content for the target object.

[0137] Step 718: Multiply the target resource consumption factor corresponding to the candidate push content by the expected resource consumption of the candidate push content for the target object to obtain the target resource consumption of the candidate push content for the target object.

[0138] Step 720: For each of the multiple target objects, based on the multiple candidate push contents corresponding to the target object and targeting the target resource consumption of the target object, select the target push content corresponding to the target object from the multiple candidate push contents corresponding to the target object, and push the target push content corresponding to the target object to the target object's terminal.

[0139] Step 722: Based on multiple historical push records within the first historical time period, update the adjustment coefficient of the first factor and return to step 706.

[0140] In this embodiment, the resource consumption factor is adjusted by the first factor adjustment coefficient and the second factor adjustment coefficient, so that the difference between the target resource consumption calculated by the adjusted resource consumption factor and the expected resource consumption is minimized, thereby reducing the impact on resource consumption.

[0141] The content push method provided in this application can be applied to any scenario that requires content push. Taking an advertising scenario as an example, this content push method can be used to push advertisements. Specifically, for candidate advertisements targeting a target object, the server can determine the resource consumption factors corresponding to the candidate advertisements on at least two business objectives, obtain each candidate resource consumption factor, determine the target resource consumption factor corresponding to the candidate advertisement based on each candidate resource consumption factor, and obtain the target resource consumption of the candidate advertisement for the target object based on the target resource consumption factor corresponding to the candidate advertisement and the expected resource consumption of the candidate advertisement for the target object. Based on the target resource consumption, content push processing is performed on the target object. For example, among the candidate advertisements for the target object, the candidate advertisement with the largest target resource consumption can be selected to obtain the target advertisement, and the target advertisement can be pushed to the target object's terminal. For example, the content push method provided in this application can be applied to... Figure 8 Within the framework of the push advertisement shown, Figure 8 In this process, the data production logic of each strategy factor is uniformly implemented and managed within the strategy framework, and accessed to the bidding service through a unified data path such as a message queue (MQ). The bidding service takes over the various strategy factors related to bidding and implements a personalized bidding framework to generate personalized bidding factors. The bidding service then passes the processed personalized bidding factors to the fine-tuning service for further processing. For example, it can use a remote procedure call (RPC) to pass the personalized bidding factors to the fine-tuning service, so that the fine-tuning service can calculate the target resource consumption. Taking the expected resource consumption as eCPM as an example, ... Figure 9As shown, an eCPM can be calculated in the coarse ranking, which can be used as the expected resource consumption. In the fine ranking, the eCPM can be recalculated using the algorithm mechanism (i.e., the target resource consumption is obtained). The strategy factor production logic is unified into the strategy framework, and the bidding-related strategy factors are uniformly connected to the bidding service. A unified data path is used between the strategy framework and the bidding service. Figure 9 The “audience scoring” can be used to determine the initial resource consumption factor of candidate push content in terms of business objectives, and the “log aggregation” can be used to obtain historical push records.

[0142] The content delivery method proposed in this application is applied to the field of ad delivery. By selecting one factor from N personalized bidding factors to influence eCPM, the impact of multiple factors acting repeatedly on ad costs is reduced. A personalized bidding constraint method is employed, and for a single personalized bidding factor, feedback adjustment is used to calculate a real-time scaling factor based on the factor multiplied by the ad granularity, thereby correcting the personalized bidding factor and minimizing its impact on ad costs. Through a dynamic adjustment method, for ads affected by personalized bidding, feedback adjustment is used to calculate a real-time scaling factor based on the ad's current real-time cost. For ads currently experiencing significant cost overruns, the ad granularity is used to correct all personalized bidding factors, minimizing their impact on ad GMV.

[0143] In some embodiments, the server can also use a learner to select the target resource consumption factor corresponding to the candidate push content from among the candidate resource consumption factors. For example, it can utilize... Figure 10 Each learner, as shown, selects the target resource consumption factor corresponding to the candidate push content from among the candidate resource consumption factors. In the diagram, the output of the base learner serves as the input to the secondary learner, with the label remaining unchanged. The secondary learner is trained by training each base learner to generate a personalized bidding factor. The input to the secondary learner is the personalized bidding factor output by each base learner, and the output is a personalized bidding factor, i.e., the target resource consumption factor. The label is the backend target data returned by the advertiser; each industry has a unique backend target.

[0144] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0145] Based on the same inventive concept, this application also provides a content push device for implementing the content push method described above. The solution provided by this device is similar to the implementation described in the above method; therefore, the specific limitations in one or more content push device embodiments provided below can be found in the limitations of the content push method described above, and will not be repeated here.

[0146] In some embodiments, such as Figure 11 As shown, a content push device is provided, including: a candidate resource consumption factor obtaining module 1102, a target resource consumption factor determining module 1104, a target resource consumption amount obtaining module 1106, and a content push processing module 1108, wherein:

[0147] The candidate resource consumption factor acquisition module 1102 is used to determine the resource consumption factors corresponding to the candidate push content for the target object on at least two business objectives, and obtain each candidate resource consumption factor.

[0148] The target resource consumption factor determination module 1104 is used to determine the target resource consumption factor corresponding to the candidate push content based on each candidate resource consumption factor.

[0149] The target resource consumption acquisition module 1106 is used to obtain the target resource consumption of the candidate push content for the target object based on the target resource consumption factor corresponding to the candidate push content and the expected resource consumption of the candidate push content for the target object; the expected resource consumption is the expected resource consumption when the push object of the candidate push content is the target object.

[0150] The content push processing module 1108 is used to push content to target objects based on the target resource consumption.

[0151] In some embodiments, the target resource consumption factor determination module is further configured to determine the data distribution information of each candidate resource consumption factor; and based on the data distribution information, select the target resource consumption factor corresponding to the candidate push content from each candidate resource consumption factor.

[0152] In some embodiments, the target resource consumption factor determination module is further configured to determine the number of candidate resource consumption factors that are greater than a preset value among the candidate resource consumption factors, to obtain a first number; determine the number of candidate resource consumption factors that are less than a preset value among the candidate resource consumption factors, to obtain a second number; and select the target resource consumption factor corresponding to the candidate push content from among the candidate resource consumption factors based on the first number and the second number.

[0153] In some embodiments, the target resource consumption factor determination module is further configured to determine a factor screening strategy based on the first quantity and the second quantity when the first quantity and the second quantity are different; and select the target resource consumption factor corresponding to the candidate push content from each candidate resource consumption factor based on the factor screening strategy.

[0154] In some embodiments, the candidate resource consumption factor obtaining module is further configured to determine the initial resource consumption factors corresponding to the candidate push content on at least two business objectives for the target object; for each of the at least two business objectives, obtain the first factor adjustment coefficient of the business objective for the candidate push content; and adjust the initial resource consumption factor of the candidate push content on the business objective using the first factor adjustment coefficient of the business objective for the candidate push content to obtain the candidate resource consumption factor of the candidate push content on the business objective.

[0155] In some embodiments, there are multiple target objects, and each target object corresponds to multiple candidate push content. The apparatus further includes: a target push content selection module, used to select target push content corresponding to the target object from the multiple candidate push content corresponding to the target object based on the target resource consumption of the target object; a content push module, used to push the target push content corresponding to the target object to the target object's terminal for each target object; a coefficient update module, used to update the first factor adjustment coefficients of at least two business objectives for the candidate push content based on multiple historical push records within a first historical time period; and return to the step of adjusting the initial resource consumption factor of the candidate push content on the business objective using the first factor adjustment coefficients of the business objective for the candidate push content to obtain the candidate resource consumption factor of the candidate push content on the business objective.

[0156] In some embodiments, the coefficient update module is further configured to, for each of at least two business objectives, select from multiple historical push records a historical push record that records the identifier of the candidate push content and the identifier of the business objective, to obtain a reference push record corresponding to the business objective; determine a reference object corresponding to the business objective based on the object identifier recorded in each reference push record corresponding to the business objective; determine the initial resource consumption factor of the candidate push content on the business objective for the reference object; and update the first factor adjustment coefficient of the business objective for the candidate push content based on the initial resource consumption factor.

[0157] In some embodiments, the coefficient update module is further configured to obtain the first resource consumption of the reference object corresponding to the business objective for the candidate push content; the first resource consumption refers to the expected resource consumption under the preset resource consumption strategy; and update the first factor adjustment coefficient of the business objective for the candidate push content based on the initial resource consumption factor of the reference object and the first resource consumption of the reference object for the candidate push content.

[0158] In some embodiments, the candidate resource consumption factor obtaining module is further configured to, for each of at least two business objectives, obtain a first correlation degree representation value between the target object and the business objective and a second correlation degree representation value between the candidate push content and the business objective; and, based on the first correlation degree representation value and the second correlation degree representation value, determine the initial resource consumption factor of the candidate push content on the business objective for the target object.

[0159] In some embodiments, the target resource consumption obtaining module is further configured to obtain the actual total resource consumption and the expected total resource consumption of the candidate push content in the second historical time period; adjust the target resource consumption factor based on the actual total resource consumption and the expected total resource consumption to obtain the adjusted target resource consumption factor; and obtain the target resource consumption of the candidate push content for the target object based on the adjusted target resource consumption factor and the expected resource consumption of the candidate push content for the target object.

[0160] In some embodiments, the target resource consumption obtaining module is further configured to determine the second factor adjustment coefficient based on the actual total resource consumption and the expected total resource consumption; and adjust the target resource consumption factor using the second factor adjustment coefficient to obtain the adjusted target resource consumption factor.

[0161] In some embodiments, the target resource consumption obtaining module is further configured to determine a first threshold based on the second factor adjustment coefficient; the first threshold is negatively correlated with the second factor adjustment coefficient; when the target resource consumption factor is less than the first threshold, the target resource consumption factor is increased based on the second factor adjustment coefficient to obtain the adjusted target resource consumption factor.

[0162] Each module in the aforementioned content delivery device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.

[0163] In some embodiments, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 12 As shown, this computer device includes a processor, memory, input / output interfaces (I / O), and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores data involved in the content delivery method. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communicating with external terminals via a network connection. When the computer program is executed by the processor, it implements a content delivery method.

[0164] In some embodiments, a computer device is provided, which may be a terminal, and its internal structure diagram may be as follows: Figure 13As shown, the computer device includes a processor, memory, input / output interfaces, a communication interface, a display unit, and an input device. The processor, memory, and input / output interfaces are connected via a system bus, and the communication interface, display unit, and input device are also connected to the system bus via the input / output interfaces. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The input / output interfaces are used for exchanging information between the processor and external devices. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, NFC (Near Field Communication), or other technologies. When the computer program is executed by the processor, it implements a content delivery method. The display unit of the computer device is used to form a visually visible image. It can be a display screen, a projection device, or a virtual reality imaging device. The display screen can be an LCD screen or an e-ink screen. The input device of the computer device can be a touch layer covering the display screen, or buttons, trackballs, or touchpads set on the casing of the computer device, or external keyboards, touchpads, or mice, etc.

[0165] Those skilled in the art will understand that Figure 12 and Figure 13 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0166] In some embodiments, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described content push method.

[0167] In some embodiments, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps in the content push method described above.

[0168] In some embodiments, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the content push method described above.

[0169] 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, data stored, data displayed, 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 the relevant data shall comply with the relevant laws, regulations and standards of the relevant countries and regions.

[0170] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0171] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0172] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A content push method, characterized in that, The method includes: For candidate push content for the target object, determine the resource consumption factor corresponding to the candidate push content on at least two business objectives respectively, and obtain each candidate resource consumption factor; Determine the data distribution information of each candidate resource consumption factor; determine the number of candidate resource consumption factors that are greater than a preset value among each candidate resource consumption factor to obtain a first number; determine the number of candidate resource consumption factors that are less than the preset value among each candidate resource consumption factor to obtain a second number; When the first quantity equals the second quantity, the preset resource consumption factor is used as the target resource consumption factor corresponding to the candidate push content. When the first quantity is greater than the second quantity, the largest candidate resource consumption factor is selected as the target resource consumption factor corresponding to the candidate push content. When the first quantity is less than the second quantity, the smallest candidate resource consumption factor is selected as the target resource consumption factor corresponding to the candidate push content. Based on the target resource consumption factor corresponding to the candidate push content and the expected resource consumption of the candidate push content for the target object, the target resource consumption of the candidate push content for the target object is obtained; the expected resource consumption is the expected resource consumption when the push object of the candidate push content is the target object. Content is pushed to the target object based on the target resource consumption.

2. The method according to claim 1, characterized in that, The step of determining the resource consumption factors corresponding to the candidate push content for at least two business objectives, and obtaining each candidate resource consumption factor, includes: Determine the initial resource consumption factors corresponding to the candidate push content for the target object on each of the at least two business objectives; For each of the at least two business objectives, obtain the first factor adjustment coefficient of the business objective for the candidate push content; By using the first factor adjustment coefficient of the business objective for the candidate push content, the initial resource consumption factor of the candidate push content on the business objective is adjusted to obtain the candidate resource consumption factor of the candidate push content on the business objective.

3. The method according to claim 2, characterized in that, The target objects are multiple, and each target object corresponds to multiple candidate push content; the method further includes: For each target object, based on the multiple candidate push contents corresponding to the target object, and targeting the target resource consumption of the target object, the target push content corresponding to the target object is selected from the multiple candidate push contents corresponding to the target object. For each target object, push the target content corresponding to the target object to the target object's terminal; Based on multiple historical push records within the first historical time period, update the first factor adjustment coefficients of the at least two business objectives for the candidate push content; The step involves adjusting the initial resource consumption factor of the candidate push content on the business objective using the first factor adjustment coefficient of the business objective, thereby obtaining the candidate resource consumption factor of the candidate push content on the business objective.

4. The method according to claim 3, characterized in that, The step of updating the first factor adjustment coefficients for the candidate push content for the at least two business objectives based on multiple historical push records within the first historical time period includes: For each of the at least two business objectives, select the historical push record containing the identifier of the candidate push content and the identifier of the business objective from multiple historical push records to obtain the reference push record corresponding to the business objective; Based on the object identifier recorded in each reference push record corresponding to the business objective, the reference object corresponding to the business objective is determined; Determine the initial resource consumption factor of the candidate push content on the business objective for the reference object, and update the first factor adjustment coefficient of the business objective for the candidate push content based on the initial resource consumption factor.

5. The method according to claim 4, characterized in that, The step of updating the first factor adjustment coefficient for the candidate push content based on the initial resource consumption factor includes: Obtain the first resource consumption of the reference object corresponding to the business objective for the candidate push content; the first resource consumption refers to the expected resource consumption under the preset resource consumption strategy; Based on the initial resource consumption factor of the reference object and the first resource consumption of the reference object for the candidate push content, the first factor adjustment coefficient of the business objective for the candidate push content is updated.

6. The method according to claim 2, characterized in that, The determination of the initial resource consumption factors corresponding to the candidate push content for the target object on each of the at least two business objectives includes: For each of the at least two business objectives, obtain a first correlation degree representation value between the target object and the business objective, and a second correlation degree representation value between the candidate push content and the business objective; Based on the first correlation degree characterization value and the second correlation degree characterization value, the initial resource consumption factor of the candidate push content on the business objective is determined for the target object.

7. The method according to any one of claims 1 to 6, characterized in that, The step of obtaining the target resource consumption of the candidate push content for the target object based on the target resource consumption factor corresponding to the candidate push content and the expected resource consumption of the candidate push content for the target object includes: Obtain the actual total resource consumption and the expected total resource consumption of the candidate push content within the second historical time period; The target resource consumption factor is adjusted based on the actual total resource consumption and the expected total resource consumption to obtain the adjusted target resource consumption factor. Based on the adjusted target resource consumption factor and the expected resource consumption of the candidate push content for the target object, the target resource consumption of the candidate push content for the target object is obtained.

8. The method according to claim 7, characterized in that, The adjustment of the target resource consumption factor based on the actual total resource consumption and the expected total resource consumption, resulting in the adjusted target resource consumption factor, includes: Based on the actual total resource consumption and the expected total resource consumption, determine the adjustment coefficient for the second factor; The target resource consumption factor is adjusted using the second factor adjustment coefficient to obtain the adjusted target resource consumption factor.

9. The method according to claim 8, characterized in that, The adjustment of the target resource consumption factor using the second factor adjustment coefficient to obtain the adjusted target resource consumption factor includes: A first threshold is determined based on the adjustment coefficient of the second factor; the first threshold is negatively correlated with the adjustment coefficient of the second factor. If the target resource consumption factor is less than the first threshold, the target resource consumption factor is increased based on the second factor adjustment coefficient to obtain the adjusted target resource consumption factor.

10. A content push device, characterized in that, The device includes: The candidate resource consumption factor acquisition module is used to determine the resource consumption factors corresponding to the candidate push content of the target object on at least two business objectives, and obtain each candidate resource consumption factor. The target resource consumption factor determination module is used to determine the data distribution information of each candidate resource consumption factor; determine the number of candidate resource consumption factors that are greater than a preset value to obtain a first number; determine the number of candidate resource consumption factors that are less than the preset value to obtain a second number; when the first number equals the second number, the preset resource consumption factor is used as the target resource consumption factor corresponding to the candidate push content; when the first number is greater than the second number, the largest candidate resource consumption factor is selected as the target resource consumption factor corresponding to the candidate push content; when the first number is less than the second number, the smallest candidate resource consumption factor is selected as the target resource consumption factor corresponding to the candidate push content. The target resource consumption acquisition module is used to obtain the target resource consumption of the candidate push content for the target object based on the target resource consumption factor corresponding to the candidate push content and the expected resource consumption of the candidate push content for the target object; the expected resource consumption is the expected resource consumption when the push object of the candidate push content is the target object. The content push processing module is used to push content to the target object based on the target resource consumption.

11. The apparatus according to claim 10, characterized in that, The candidate resource consumption factor obtaining module is further configured to determine the initial resource consumption factor corresponding to the candidate push content for the target object on at least two business objectives; and for each of the at least two business objectives, to obtain the first factor adjustment coefficient of the business objective for the candidate push content. By using the first factor adjustment coefficient of the business objective for the candidate push content, the initial resource consumption factor of the candidate push content on the business objective is adjusted to obtain the candidate resource consumption factor of the candidate push content on the business objective.

12. The apparatus according to claim 11, characterized in that, The device further includes: a target push content selection module, used to select target push content corresponding to the target object from the multiple candidate push content corresponding to the target object based on the target resource consumption of the target object; a content push module, used to push the target push content corresponding to the target object to the target object's terminal for each target object; a coefficient update module, used to update the first factor adjustment coefficients of at least two business objectives for the candidate push content based on multiple historical push records within a first historical time period; and a step of adjusting the initial resource consumption factor of the candidate push content on the business objective using the first factor adjustment coefficients of the business objective for the candidate push content to obtain the candidate resource consumption factor of the candidate push content on the business objective.

13. The apparatus according to claim 12, characterized in that, The coefficient update module is further configured to, for each of at least two business objectives, select from multiple historical push records a historical push record that records the identifier of the candidate push content and the identifier of the business objective, to obtain a reference push record corresponding to the business objective; and determine a reference object corresponding to the business objective based on the object identifier recorded in each reference push record corresponding to the business objective. Determine the initial resource consumption factor of the candidate push content for the reference object in terms of business objectives, and update the first factor adjustment coefficient of the business objectives for the candidate push content based on the initial resource consumption factor.

14. The apparatus according to claim 13, characterized in that, The coefficient update module is further used to obtain the first resource consumption of the reference object corresponding to the business objective for the candidate push content; the first resource consumption refers to the expected resource consumption under the preset resource consumption strategy; based on the initial resource consumption factor of the reference object and the first resource consumption of the reference object for the candidate push content, the first factor adjustment coefficient of the business objective for the candidate push content is updated.

15. The apparatus according to claim 10, characterized in that, The candidate resource consumption factor obtaining module is further configured to obtain, for each of at least two business objectives, a first correlation degree representation value between the target object and the business objective and a second correlation degree representation value between the candidate push content and the business objective; Based on the first correlation degree characterization value and the second correlation degree characterization value, the initial resource consumption factor of the candidate push content in terms of business objectives is determined for the target object.

16. The apparatus according to any one of claims 10 to 15, characterized in that, The target resource consumption acquisition module is also used to obtain the actual total resource consumption and the expected total resource consumption of the candidate push content in the second historical time period; and to adjust the target resource consumption factor based on the actual total resource consumption and the expected total resource consumption to obtain the adjusted target resource consumption factor. Based on the adjusted target resource consumption factor and the expected resource consumption of the candidate push content for the target audience, the target resource consumption of the candidate push content for the target audience is obtained.

17. The apparatus according to claim 16, characterized in that, The target resource consumption acquisition module is further configured to determine the second factor adjustment coefficient based on the actual total resource consumption and the expected total resource consumption; and to adjust the target resource consumption factor using the second factor adjustment coefficient to obtain the adjusted target resource consumption factor.

18. The apparatus according to claim 17, characterized in that, The target resource consumption obtaining module is further configured to determine a first threshold based on the second factor adjustment coefficient; the first threshold is negatively correlated with the second factor adjustment coefficient; when the target resource consumption factor is less than the first threshold, the target resource consumption factor is increased based on the second factor adjustment coefficient to obtain the adjusted target resource consumption factor.

19. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 9.

20. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 9.

21. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 9.

Images