A multi-scene-oriented AI advertisement delivery optimization method and system

Abstract

The application relates to the technical field of advertisement putting optimization, and discloses an AI advertisement putting optimization method and system for multiple scenes, which comprises the following steps: establishing an advertisement content library, wherein all advertisement contents required to be put are included in the advertisement content library; classifying advertisement types in the advertisement content library in combination with AI technology; setting a putting frequency threshold; and sequentially putting the advertisements in the advertisement content library in different putting channels. According to the application, the establishment of the advertisement content library and the automatic classification of the advertisements by using the AI technology can dynamically distinguish high-matching and low-matching type advertisements according to putting feedback by monitoring the click rates of various types of advertisements, directly reflect the acceptance and demand degree of users in different putting scenes, effectively incline the advertisement resources to advertisements with better effects, and thus improve the overall click rate and conversion rate.

Description

Technical Field

[0001] This invention relates to the field of advertising optimization technology, specifically to an AI advertising optimization method and system for multiple scenarios. Background Technology

[0002] Multi-scenario advertising refers to displaying and promoting advertisements on multiple different media, platforms, or user touchpoints, rather than being limited to a single channel. Examples include search engines, social media, video platforms, mobile applications, and outdoor media. Different channels have their own user groups and communication methods, enabling them to reach a wider audience.

[0003] Traditional multi-scenario advertising methods simply rotate the advertising content to be placed on different channels. However, the user groups, behaviors, and needs of each channel are quite different. Simply rotating different advertising content may not be able to meet the specific needs of users in different scenarios. This lack of real-time feedback and optimization mechanisms for each scenario can lead to poor advertising performance on some channels, resulting in unsatisfactory overall advertising returns. Summary of the Invention

[0004] To address the problems existing in the prior art, the present invention aims to provide an AI advertising delivery optimization method and system for multiple scenarios, which can utilize AI technology to deliver targeted advertising content that better meets the needs of users on different channels, thereby improving ad click-through rates and conversion rates.

[0005] To achieve the above objectives, the present invention provides the following technical solution: an AI advertising optimization method for multiple scenarios, the method comprising the following steps:

[0006] Establish an advertising content library, which includes all advertising content to be delivered. Combine AI technology to classify the types of ads in the advertising content library, set a threshold for the number of times ads are delivered, and deliver the ads in the advertising content library in different delivery channels in sequence until the number of times ads are delivered in the corresponding delivery channels reaches the preset threshold for the number of times ads are delivered.

[0007] Get the click-through rate (CTR) of different types of ads in the delivery channels, set CTR thresholds, and mark the ad type as a high-match ad when the CTR of the corresponding ad type is greater than or equal to the CTR threshold; mark the ad type as a low-match ad when the CTR of the corresponding ad type is less than the CTR threshold.

[0008] Determine the number of high-match and low-match ads in the advertising channel. If all ads in the advertising channel are either high-match or low-match ads, the sequential advertising method will still be executed. If both high-match and low-match ads exist in the advertising channel, the advertising optimization strategy will be executed.

[0009] The campaign optimization strategy includes obtaining the number of high-match type ads, combining the ad frequency with the number of high-match type ads to obtain the number of campaign adjustments, reducing the frequency of all low-match type ads in the campaign channel based on the number of campaign adjustments, and increasing the frequency of all high-match type ads based on the number of campaign adjustments.

[0010] The specific way to obtain the number of ad placement adjustments is to divide the ad placement frequency threshold by the number of ad types to obtain the ad placement frequency, and then combine the ad placement frequency Tb with the number of highly matched type ads Gs to obtain the number of ad placement adjustments Tc = Tb ÷ (1 + Gs).

[0011] The specific method for adjusting the frequency of high-match and low-match type ads based on the number of ad adjustments is as follows: For each low-match type ad in the ad placement channel, the number of ad adjustments will be used instead of the ad placement frequency for subsequent ad placements; the frequency increase obtained by multiplying the number of ad adjustments by the number of low-match type ads is summed with the ad placement frequency to obtain the ad boost frequency; and each high-match type ad will use the ad boost frequency instead of the ad placement frequency for subsequent ad placements.

[0012] In some implementations, the frequency of high-match type ads is increased based on the number of ad placement adjustments. The click-through rate (CTR) after the increase in the frequency of high-match type ads is obtained, and then divided by the CTR before the increase in the frequency of ads of that type to obtain the change in CTR. A threshold for the change in CTR is set, and the change in CTR is compared with the threshold for the change in CTR. A corresponding response is made based on the comparison result.

[0013] In some implementations, if the change in click-through rate is greater than or equal to the change threshold, the campaign volume will continue to be increased according to the campaign optimization strategy without any additional adjustments; if the change in click-through rate is less than the change threshold, the category optimization strategy will be implemented.

[0014] In some implementations, the classification optimization strategy includes obtaining the click-through rate (CTR) of all low-match type ads in the delivery channel, setting a secondary CTR threshold that is less than the primary CTR threshold, comparing the highest CTR among the low-match type ads with the secondary CTR threshold, and if the highest CTR among the low-match type ads is greater than or equal to the secondary CTR threshold, then swapping the low-match type ad with the corresponding high-match type ad; if the highest CTR among the low-match type ads is less than the secondary CTR threshold, then maintaining the classification of the corresponding high-match type ad without further adjustment.

[0015] In some implementations, when the highest click-through rate (CTR) of a low-match type ad is greater than or equal to a secondary CTR threshold, the conversion effect of that low-match type ad is obtained, and the conversion effect of a high-match type ad whose corresponding CTR change is less than a change threshold is also obtained. The difference between the two conversion effects is then determined. If the conversion effect of the high-match type ad is greater than that of the low-match type ad, the two types of ads are not swapped. If the conversion effect of the high-match type ad is less than or equal to that of the low-match type ad, the determination to swap the two types of ads is maintained.

[0016] In some implementations, the specific way to obtain the advertising conversion effect is to obtain the number of times the corresponding advertisement is clicked and the number of times the corresponding advertisement achieves its advertising goal, and then obtain the corresponding advertising conversion effect by dividing the number of times the advertising goal is achieved by the number of times the advertisement is clicked.

[0017] This invention also provides the following technical solution: an AI advertising delivery optimization system for multiple scenarios, comprising:

[0018] The data collection module includes establishing an advertising content library containing all advertising content to be delivered, and using AI technology to classify the types of advertisements in the advertising content library, setting a threshold for the number of times advertisements are delivered, and delivering the advertisements in the advertising content library sequentially on different delivery channels until the number of times advertisements are delivered on the corresponding delivery channels reaches the preset threshold for the number of times advertisements are delivered.

[0019] The category classification module includes obtaining the click-through rate (CTR) of different types of ads in the delivery channel, setting a CTR threshold, and marking the ad type as a high-match ad when the CTR of the corresponding ad type is greater than or equal to the CTR threshold; and marking the ad type as a low-match ad when the CTR of the corresponding ad type is less than the CTR threshold.

[0020] The strategy judgment module includes determining the number of high-match type ads and low-match type ads in the delivery channel. When all the ads in the delivery channel are high-match type ads or low-match type ads, the sequential delivery method is still executed. When both high-match type ads and low-match type ads exist in the delivery channel, the delivery optimization strategy is executed.

[0021] The optimization execution module includes executing a delivery optimization strategy, which includes obtaining the number of high-match type ads, combining the ad delivery frequency with the number of high-match type ads to obtain the number of delivery adjustment times, reducing the delivery frequency of all low-match type ads in the delivery channel based on the number of delivery adjustment times, and increasing the delivery frequency of all high-match type ads based on the number of delivery adjustment times.

[0022] The present invention further provides a computer-readable storage medium storing a computer program, which is executed by a processor to implement the above-described AI advertising optimization method for multiple scenarios.

[0023] The technical solution provided by this invention has the following advantages compared with the prior art:

[0024] Firstly, this invention establishes an advertising content library and uses AI technology to automatically classify advertisements. By monitoring the click-through rate of each type of advertisement, it can dynamically distinguish between high-match and low-match advertisements based on the delivery feedback. This directly reflects the user's acceptance and demand in different delivery scenarios, effectively tilting advertising resources towards advertisements with better performance, thereby improving the overall click-through rate and conversion rate and achieving higher delivery efficiency. This method can flexibly adapt advertising strategies for different scenarios and user groups, thus achieving good advertising results in multiple scenarios.

[0025] Secondly, by monitoring the changes in click-through rate of high-match ads after the frequency of ad placement increases, this invention can promptly identify and address user fatigue. When it is found that the click-through rate of high-match ads has declined significantly after the frequency of ad placement increases, a classification optimization strategy is used to introduce low-match ads with near-ideal performance and exchange types to ensure that the ad content always maintains a high degree of matching. This can make full use of high-quality ad resources and prevent users from resisting repetitive or over-exposed content.

[0026] Thirdly, this invention can reflect the extent to which users achieve preset goals after clicking on an ad through conversion results. Introducing dual indicators can more accurately evaluate the actual effect and commercial value of an ad, avoiding misjudgments caused by relying solely on click-through rate. This dual-indicator judgment strategy helps to filter out ad content that is both attractive to users and can bring actual conversions, thereby improving the overall conversion rate and return on investment of ad placement, enhancing the comprehensiveness and accuracy of ad evaluation, and effectively preventing frequent adjustments due to fluctuations in a single click-through rate. Attached Figure Description

[0027] Figure 1 This is a flowchart illustrating an AI advertising delivery optimization method for multiple scenarios according to the present invention.

[0028] Figure 2 This is a schematic diagram of a multi-scenario AI advertising delivery optimization system according to the present invention. Detailed Implementation

[0029] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0030] It is understood that the term "a" should be understood as "at least one" or "one or more", that is, in one embodiment, the number of an element can be one, while in another embodiment, the number of the element can be multiple, and the term "a" should not be understood as a limitation on the number.

[0031] This invention provides an AI-powered advertising optimization method for multiple scenarios, such as... Figure 1 As shown, the method includes the following steps:

[0032] Step 1: Establish an advertising content library. The advertising content library includes all advertising content to be placed. Combine AI technology to classify the types of advertisements in the advertising content library. For example, use natural language processing (NLP) and computer vision technology to automatically classify, extract tags and extract content features of advertising materials to distinguish and classify different types of advertisements. Set a threshold for the number of times to place advertisements. Place the advertisements in the advertising content library in different placement channels in sequence until the number of times to place advertisements in the corresponding placement channels reaches the preset threshold.

[0033] Step 2: Obtain the click-through rate (CTR) for different types of ads in the advertising channels, set CTR thresholds, and compare the CTR of each ad type with the CTR threshold. When the CTR of a corresponding ad type is greater than or equal to the CTR threshold, it indicates that this type of ad can achieve an ideal CTR when advertised through this channel, and users in this advertising scenario have a high interest in this type of ad. This type of ad is marked as a high-match ad. When the CTR of a corresponding ad type is less than the CTR threshold, it indicates that this type of ad has an unsatisfactory CTR when advertised through this channel, and users in this advertising scenario have insufficient interest in this type of ad. This type of ad is marked as a low-match ad.

[0034] Step 3: Determine the number of high-match and low-match ads in the delivery channel. If all ads in the delivery channel are either high-match or low-match ads, it indicates that all types of ads in the ad content library are receiving uniformly high or low click-through rates in this delivery channel. In this case, continue with the sequential delivery method without any additional adjustments. If both high-match and low-match ads exist in the delivery channel, it indicates that there are differences in user acceptance and demand among different types of ads in the ad content library in this delivery channel. In this case, implement a delivery optimization strategy.

[0035] Step four, implementing the campaign optimization strategy, includes obtaining the number of high-match type ads, and combining the ad frequency with the number of high-match type ads to obtain the number of campaign adjustments. The campaign frequency of all low-match type ads in the campaign channel is reduced based on the number of campaign adjustments, while the campaign frequency of all high-match type ads is increased based on the number of campaign adjustments, in order to increase the campaign frequency of high-match type ads.

[0036] In step one, ads in the ad content library are sequentially delivered across different delivery channels until the number of deliveries reaches the delivery threshold. Therefore, the specific way to obtain the ad delivery frequency is to divide the delivery threshold by the number of ad types. The resulting ad delivery frequency represents the number of times each ad type is delivered after the delivery threshold is reached in step one. By combining the ad delivery frequency Tb with the number of high-match ad types Gs, the delivery adjustment number Tc = Tb ÷ (1 + Gs) can be obtained. After obtaining the specific delivery adjustment number, each low-match ad type in the delivery channel will use the delivery adjustment number instead of the delivery frequency to perform subsequent ad delivery. Each high-match ad type in the delivery channel will have its original delivery frequency increased by adding the delivery adjustment number multiplied by the number of low-match ad types. The two are combined to obtain the delivery boost frequency, which is then used to replace the original delivery frequency for subsequent ad delivery.

[0037] For example, in an ad content library, an AI-based application categorizes the ads placed on various channels into five types: Type A, Type B, Type C, Type D, and Type E. The threshold for ad placement is set at 100 times. Under a sequential placement rule, each type of ad is placed 20 times (i.e., the ad placement frequency is 20 times). The click-through rate (CTR) threshold is set at 10%. If Type A, Type B, Type C, Type D, and Type E ads are clicked 0, 2, 3, 5, and 0 times respectively, the CTRs are 0%, 10%, 15%, 25%, and 0%. Then, Type B, Type C, and Type D ads are marked as high-match ads, while Type A and Type E ads are marked as low-match ads. An optimization strategy is then implemented, focusing on ad placement frequency at 20 times and targeting high-match ads. With 3 ads, the number of adjustment times is 20 ÷ (1 + 3) = 5. Therefore, when executing ad campaigns, type A and type E ads (low-match ads) will each be run only 5 times out of every 100 ad campaigns, while type B, type C, and type D ads (high-match ads) will each be run 20 + 5 × 2 = 30 times out of every 100 ad campaigns. This allows high-match ads to have a higher reach and exposure among users in this scenario. The above method, by integrating AI technology, not only achieves intelligent classification of ad content but also builds a dynamic adjustment mechanism based on real-time feedback. This enables different ad types to achieve accurate user matching and resource allocation in multi-channel campaigns, thereby effectively improving ad exposure, click-through rate, and final conversion effect, thus optimizing overall ad performance and revenue.

[0038] It should be noted that if the click-through rate of each type of ad in the second step is 0, it means that all ads in the ad content library will not achieve effective conversion in this ad channel. This will indicate that the ad channel is not compatible with the ad content library, and it is advisable to stop ad placement in this scenario to avoid losses caused by ineffective placement.

[0039] After increasing the frequency of all high-match ad types based on the number of ad placement adjustments, the click-through rate (CTR) after the increase in the frequency of high-match ad types is obtained. This CTR is then divided by the CTR before the increase to obtain the change in CTR. A threshold for this change is set, and the change in CTR is compared with the threshold. Based on the comparison results, the following actions are taken: If the change in CTR is greater than or equal to the threshold, it indicates that the increase in the frequency of high-match ad placements has not resulted in a significant drop in the CTR for that ad type, meaning that user interest and interaction with that ad type remain at a high level. In this case, the strategy of increasing the frequency of ad placements continues without further adjustments. If the change in CTR is less than the threshold, it indicates that the increase in the frequency of high-match ad placements has resulted in a significant drop in the CTR for that ad type. Even if the CTR for that ad type can still maintain the level of the CTR threshold, it means that the ad content has reached a user saturation point with the increased impressions, meaning that users are beginning to experience fatigue with this type of ad. Continuing with a large number of placements may actually lead to user aversion. In this case, a category optimization strategy is implemented.

[0040] The classification optimization strategy includes obtaining the click-through rate (CTR) of all low-match type ads in the advertising channels and setting a secondary CTR threshold lower than the primary CTR threshold. The highest CTR among the low-match type ads is compared with this secondary CTR threshold. If the highest CTR among the low-match type ads is greater than or equal to the secondary CTR threshold, it indicates that there are low-match type ads in the current ad content library with CTRs close to the primary CTR threshold. In this case, the low-match type ad is swapped with its corresponding high-match type ad. If the highest CTR among the low-match type ads is less than the secondary CTR threshold, it indicates that the CTRs of all low-match type ads in the current ad content library are unsatisfactory. In this case, the classification of the corresponding high-match type ad is maintained without further adjustment. This is because, even when there are no other low-match type ads with ideal CTRs in the ad content library, and the CTR of this high-match type ad still reaches the CTR threshold, maintaining the judgment of the advertising optimization strategy ensures that the advertising strategy will not be rashly adjusted due to slight data fluctuations. This avoids instability in subsequent ad conversion rates due to strategy fluctuations when the current ad performance is satisfactory.

[0041] For example, taking the above embodiment as an example, after implementing the delivery optimization strategy, the delivery frequency of type D ads is increased from 20 times out of every 100 ad deliveries to 30 times. When the number of subsequent ad deliveries reaches the delivery threshold (100 times) again, type D ads are clicked 5 times, with a click-through rate (CTR) of 17%. Since its CTR before the increase was 25%, the change in CTR is 0.68. The threshold for the change is set at 0.7. Since the change in CTR is less than the threshold, a classification optimization strategy should be implemented. The highest CTR among low-match ads is set to type A ads, which have a CTR of 9.5%. The secondary threshold for CTR is set at 8%. Since the CTR of type A ads exceeds the secondary threshold, it means that although its CTR has not reached the threshold, it is very close. Therefore, type A ads are replaced with high-match ads, and type D ads are replaced with low-match ads. By swapping categories, low-match ads that were originally performing well can be upgraded to high-match ads, giving them a higher frequency of delivery and priority in display resources. This also avoids misclassification of ad categories due to data sparsity in the early stages. At the same time, ads that were originally in the high-match category but have recently declined in performance can be downgraded to low-match ads, thereby reducing their display opportunities and mitigating potential resource waste and user fatigue risks.

[0042] Simultaneously, when the highest click-through rate (CTR) of a low-match type ad is greater than or equal to the secondary CTR threshold, the conversion effect of that low-match type ad should also be obtained, along with the conversion effect of a high-match type ad whose corresponding CTR change is less than the change threshold. The difference between these two conversion effects should be determined. If the conversion effect of the high-match type ad is greater than that of the low-match type ad, then the two types of ads should not be swapped. If the conversion effect of the high-match type ad is less than or equal to that of the low-match type ad, then the determination to swap the two types of ads should be maintained.

[0043] The specific method for obtaining ad conversion results is to obtain the number of times the corresponding ad was clicked and the number of times the ad's advertising objective was achieved. The advertising objective refers to the preset action that the user is required to complete after clicking the ad, such as a payment action, registration action, or download action. The corresponding ad conversion result is obtained by dividing the number of times the advertising objective was achieved by the number of times the ad was clicked. Taking the above embodiment as an example, if type D ad was clicked 5 times and the advertising objective was achieved 2 times, the conversion result of type D ad can be calculated as 40%. If type A ad conversion result is set to 60%, since type A ad, which belongs to a low-match ad category, has a higher conversion result than type D ad, the judgment of swapping the two ad types is maintained. This design, in addition to focusing on click-through rate (CTR), also measures the effectiveness of ads in achieving expected goals (such as payments, registrations, and downloads) through conversion performance. This avoids making decisions solely based on CTR, because when swapping high-match and low-match ads, relying solely on CTR fluctuations could lead to hasty adjustments. Introducing conversion performance as a supplementary indicator helps identify which type of ad, while generating clicks, truly drives users to complete their desired actions. If high-match ads have better conversion performance, it indicates that their overall quality is superior, and even if the CTR decreases, the category should not be easily swapped, thus avoiding frequent adjustments due to minor fluctuations.

[0044] In summary, this invention aims to design an AI-powered advertising optimization method for multiple scenarios. Addressing the issue that traditional ad content rotation may not effectively meet user needs across different scenarios, this invention establishes an ad content library, utilizes AI technology for automatic ad classification, and monitors the click-through rate (CTR) of each ad type. Based on campaign feedback, it dynamically distinguishes between high-match and low-match ads, directly reflecting user acceptance and demand in different scenarios. This effectively directs ad resources towards higher-performing ads, thereby improving overall CTR and conversion rates and achieving higher campaign efficiency. This method can flexibly adapt advertising strategies to different scenarios and user groups, achieving good advertising results across multiple scenarios. By monitoring the CTR changes of high-match ads after increasing campaign frequency, it promptly identifies and addresses user fatigue. When a significant decline in the CTR of high-match ads is observed after increasing campaign frequency, a classification optimization strategy introduces low-match ads with near-ideal performance for type swapping, ensuring that ad content maintains a high degree of relevance. This approach fully utilizes high-quality ad resources while preventing user resistance to repetitive or overly exposed content. Conversion performance reflects how well users achieve their preset goals after clicking on an ad. Introducing dual metrics allows for a more accurate assessment of the actual effectiveness and commercial value of an ad, avoiding misjudgments caused by relying solely on click-through rate. This dual-metric judgment strategy helps to filter out ad content that is both attractive to users and can bring actual conversions, thereby improving the overall conversion rate and return on investment of ad campaigns. It also enhances the comprehensiveness and accuracy of ad evaluation and effectively prevents frequent adjustments due to fluctuations in a single click-through rate.

[0045] This invention provides an AI-powered advertising delivery optimization system for multiple scenarios, such as... Figure 2 As shown, it includes:

[0046] The data collection module includes establishing an advertising content library containing all advertising content to be delivered, and using AI technology to classify the types of advertisements in the advertising content library, setting a threshold for the number of times advertisements are delivered, and delivering the advertisements in the advertising content library sequentially on different delivery channels until the number of times advertisements are delivered on the corresponding delivery channels reaches the preset threshold for the number of times advertisements are delivered.

[0047] The category classification module includes obtaining the click-through rate (CTR) of different types of ads in the delivery channel, setting a CTR threshold, and marking the ad type as a high-match ad when the CTR of the corresponding ad type is greater than or equal to the CTR threshold; and marking the ad type as a low-match ad when the CTR of the corresponding ad type is less than the CTR threshold.

[0048] The strategy judgment module includes determining the number of high-match type ads and low-match type ads in the delivery channel. When all the ads in the delivery channel are high-match type ads or low-match type ads, the sequential delivery method is still executed. When both high-match type ads and low-match type ads exist in the delivery channel, the delivery optimization strategy is executed.

[0049] The optimization execution module includes executing a delivery optimization strategy, which includes obtaining the number of high-match type ads, combining the ad delivery frequency with the number of high-match type ads to obtain the number of delivery adjustment times, reducing the delivery frequency of all low-match type ads in the delivery channel based on the number of delivery adjustment times, and increasing the delivery frequency of all high-match type ads based on the number of delivery adjustment times.

[0050] The processes described above with reference to the flowcharts in the embodiments disclosed in this invention can be implemented as computer software programs. Embodiments of this invention include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication component, and / or installed from a removable medium. When the computer program is executed by a central processing unit, it performs the functions defined in the methods of this application. It should be noted that the computer-readable medium described above in this application can be a computer-readable signal medium or a computer-readable storage medium, or any combination of the two. A computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection having one or more conductor segments, a portable computer disk, a hard disk, a random access memory, a read-only memory, an erasable programmable read-only memory, an optical fiber, a portable compact disk read-only memory, an optical storage device, a magnetic storage device, or any suitable combination thereof. In this application, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in connection with an instruction execution system, apparatus, or device. In this application, a computer-readable signal medium can include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium can also be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wireless segments, wire segments, optical cables, RF, etc., or any suitable combination thereof.

[0051] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0052] Those skilled in the art should understand that the above description is only a specific embodiment of this application, but the protection scope of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the protection scope of this application.

Claims

1. A method for optimizing AI advertising delivery across multiple scenarios, characterized in that, The method includes the following steps: Establish an advertising content library, which includes all advertising content to be delivered. Combine AI technology to classify the types of ads in the advertising content library, set a threshold for the number of times ads are delivered, and deliver the ads in the advertising content library in different delivery channels in sequence until the number of times ads are delivered in the corresponding delivery channels reaches the preset threshold for the number of times ads are delivered. Get the click-through rate (CTR) of different types of ads in the delivery channels, set CTR thresholds, and mark the ad type as a high-match ad when the CTR of the corresponding ad type is greater than or equal to the CTR threshold; mark the ad type as a low-match ad when the CTR of the corresponding ad type is less than the CTR threshold. Determine the number of high-match and low-match ads in the advertising channel. If all ads in the advertising channel are either high-match or low-match ads, the sequential advertising method will still be executed. If both high-match and low-match ads exist in the advertising channel, the advertising optimization strategy will be executed. The campaign optimization strategy includes obtaining the number of high-match type ads, combining the ad frequency with the number of high-match type ads to obtain the number of campaign adjustments, reducing the frequency of all low-match type ads in the campaign channel based on the number of campaign adjustments, and increasing the frequency of all high-match type ads based on the number of campaign adjustments. The specific way to obtain the number of ad placement adjustments is to divide the ad placement frequency threshold by the number of ad types to obtain the ad placement frequency, and then combine the ad placement frequency Tb with the number of highly matched type ads Gs to obtain the number of ad placement adjustments Tc = Tb ÷ (1 + Gs). The specific method for adjusting the frequency of high-match and low-match type ads based on the number of ad adjustments is as follows: For each low-match type ad in the ad placement channel, the number of ad adjustments will be used instead of the ad placement frequency for subsequent ad placements; the frequency increase obtained by multiplying the number of ad adjustments by the number of low-match type ads is summed with the ad placement frequency to obtain the ad boost frequency; and each high-match type ad will use the ad boost frequency instead of the ad placement frequency for subsequent ad placements.

2. The AI ​​advertising optimization method for multiple scenarios according to claim 1, characterized in that, After increasing the frequency of high-match type ads based on the number of ad placement adjustments, the click-through rate (CTR) after the increase in the frequency of high-match type ads is obtained. This CTR is then divided by the CTR before the increase in the frequency of ads of this type to obtain the change in CTR. A threshold for the change in CTR is set, and the change in CTR is compared with the threshold. Based on the comparison results, appropriate actions are taken.

3. The AI ​​advertising optimization method for multiple scenarios according to claim 2, characterized in that, If the change in click-through rate is greater than or equal to the change threshold, the campaign optimization strategy will continue to be implemented to increase the campaign volume without any additional adjustments; if the change in click-through rate is less than the change threshold, the category optimization strategy will be implemented.

4. The AI ​​advertising optimization method for multiple scenarios according to claim 3, characterized in that, The classification optimization strategy includes obtaining the click-through rate (CTR) of all low-match type ads in the delivery channel, setting a secondary CTR threshold that is lower than the primary CTR threshold, comparing the highest CTR among the low-match type ads with the secondary CTR threshold, and if the highest CTR among the low-match type ads is greater than or equal to the secondary CTR threshold, then swapping the low-match type ad with the corresponding high-match type ad; if the highest CTR among the low-match type ads is less than the secondary CTR threshold, then maintaining the classification of the corresponding high-match type ad without further adjustment.

5. The AI ​​advertising optimization method for multiple scenarios according to claim 4, characterized in that, When the highest click-through rate (CTR) of a low-match ad type is greater than or equal to the secondary CTR threshold, the conversion effect of that low-match ad type is obtained, and the conversion effect of the corresponding high-match ad type with a CTR change amount less than the change amount threshold is obtained. The difference between the two conversion effects is determined. If the conversion effect of the high-match ad type is greater than that of the low-match ad type, the two types of ads are not swapped. If the conversion effect of the high-match ad type is less than or equal to that of the low-match ad type, the determination to swap the two types of ads is maintained.

6. The AI ​​advertising optimization method for multiple scenarios according to claim 5, characterized in that, The specific way to obtain the advertising conversion effect is to get the number of times the corresponding advertisement was clicked and the number of times the corresponding advertisement achieved its advertising goal. The conversion effect of the corresponding advertisement is obtained by dividing the number of times the advertising goal was achieved by the number of times the advertisement was clicked.

7. An AI-powered advertising delivery optimization system for multiple scenarios, characterized in that, The AI ​​advertising delivery optimization method for multiple scenarios according to any one of claims 1-6 includes: The data collection module includes establishing an advertising content library containing all advertising content to be delivered, and using AI technology to classify the types of advertisements in the advertising content library, setting a threshold for the number of times advertisements are delivered, and delivering the advertisements in the advertising content library sequentially on different delivery channels until the number of times advertisements are delivered on the corresponding delivery channels reaches the preset threshold for the number of times advertisements are delivered. The category classification module includes obtaining the click-through rate (CTR) of different types of ads in the delivery channel, setting a CTR threshold, and marking the ad type as a high-match ad when the CTR of the corresponding ad type is greater than or equal to the CTR threshold; and marking the ad type as a low-match ad when the CTR of the corresponding ad type is less than the CTR threshold. The strategy judgment module includes determining the number of high-match type ads and low-match type ads in the delivery channel. When all the ads in the delivery channel are high-match type ads or low-match type ads, the sequential delivery method is still executed. When both high-match type ads and low-match type ads exist in the delivery channel, the delivery optimization strategy is executed. The optimization execution module includes executing a delivery optimization strategy, which includes obtaining the number of high-match type ads, combining the ad delivery frequency with the number of high-match type ads to obtain the number of delivery adjustment times, reducing the delivery frequency of all low-match type ads in the delivery channel based on the number of delivery adjustment times, and increasing the delivery frequency of all high-match type ads based on the number of delivery adjustment times.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, which is executed by a processor to implement the AI ​​advertising delivery optimization method for multiple scenarios as described in any one of claims 1-6.

Images