Advertisement putting platform based on AI dynamic creative optimization

Abstract

The invention discloses an advertisement putting platform based on AI dynamic creative optimization, and relates to the technical field of advertisement putting, and the platform comprises five modules: a dynamic creative basic network construction module deploys multi-modal nodes to construct a full-scene generation network; the creative reference index acquisition module acquires effect and adaptive data, and analyzes to obtain a quantitative index; the creative failure evaluation module evaluates whether the effect of the creative element is attenuated according to the index; the alternative creativity planning module analyzes available alternative elements and plans a scheme; and the creative emergency response scheme execution module executes the emergency scheme when the alternative elements are all attenuated. The effect data comprises a click conversion rate and the like, and the adaptive data comprises theme correlation degree and the like, and calculating indexes through normalization processing. Attenuation judgment refers to effect index decreasing amplitude and adaptive index fluctuation, alternative schemes are screened according to priorities, emergency schemes are rapidly generated, new creativity is tested, and the advertisement putting effect and efficiency are improved.

Description

Technical Field

[0001] This invention relates to the field of advertising delivery technology, and more specifically to an advertising delivery platform based on AI-driven dynamic creative optimization. Background Technology

[0002] In the traditional advertising field, the generation and optimization of advertising creatives have long relied on human experience, resulting in three major pain points: First, it is difficult to predict the decline in creative effectiveness due to user fatigue, intensified market competition, and dynamic changes in scenarios, while human monitoring is lagging behind. Second, the allocation of creative resources lacks precision, with most platforms adopting a "one-size-fits-all" resource allocation model, failing to dynamically adjust the resource ratio of core and auxiliary elements based on the audience coverage and expected conversion value of the advertisement. Third, the alternative creative response mechanism is rigid; when the main creative element fails, the selection of backup creatives relies heavily on historical case matching, without considering real-time changes in user preferences, resulting in an adaptation rate of less than 50% for alternative solutions. Therefore, there is an urgent need for an advertising platform based on AI-driven dynamic creative optimization.

[0003] Existing technology, such as the invention patent application with publication number CN119693063A, discloses an advertising optimization and delivery system and method, which includes the following steps: S10: acquiring user behavior data, including the user's historical interaction information and real-time environmental data; S20: analyzing the user's historical interaction information and real-time environmental data to identify user characteristics and user interests; S30: dynamically generating optimal advertising content and format that match the user characteristics and user interests based on a combined algorithm of convolutional neural network and generative adversarial network; S40: delivering the optimal advertising content and format in real time on the target advertising platform. The method of this invention, by acquiring and analyzing user behavior data, dynamically generates personalized advertising content and format, and delivers it in real time on the target platform, which helps to improve the accuracy and effectiveness of advertising.

[0004] Regarding the above solutions, the applicant of this invention has found that the above technology has at least the following technical problems: 1. Existing advertising placement technologies have significant shortcomings in the real-time performance and accuracy of dynamic creative optimization. On the one hand, the monitoring of effect decay relies excessively on manual review or simple threshold triggers, which cannot capture the non-linear downward trend of creative element effect indicators—for example, the click-through rate of an e-commerce ad may suddenly drop from 5% to 2%—nor can it correlate the impact of changes in user scenarios on creative adaptability. Optimization is often only initiated when core indicators drop by more than 30%, resulting in invalid exposure costs accounting for more than 30%. On the other hand, the resource allocation mechanism is rigid. Most platforms use a fixed ratio to allocate creative production resources without dynamically adjusting based on the audience size and expected conversion value of the ad. For example, a large amount of resources are invested in auxiliary copy with low conversion potential, while resources for optimizing core product images are insufficient, resulting in more than 40% waste of resources.

[0005] 2. Existing technologies for replacing creative systems also have significant shortcomings in adaptability and response speed. Backup creative libraries largely rely on static storage of historical cases, matching only basic tags such as product categories during screening, without considering real-time user preferences and campaign scenario characteristics. This results in a match rate of less than 50% between the alternative and the target audience. For example, a beverage brand used images from a winter scene in its summer ad creative replacement, resulting in only an 8% increase in conversion rate, far below expectations. More importantly, when the main creative and all backup creatives simultaneously decline, the lack of an automated mechanism for generating new creatives necessitates manual intervention throughout the planning, production, and review process, with response times exceeding 24 hours. This severely disrupts the advertising campaign rhythm, potentially causing millions in revenue losses, especially during critical periods such as major promotional events.

[0006] 3. The shortcomings of existing technologies in cross-channel collaboration and quantification of metrics further restrict optimization effectiveness. Inconsistent creative format requirements and data interface standards across different advertising channels necessitate manual adjustments to the size, duration, and interaction style of the same creative for each channel, resulting in low adaptation efficiency and a high risk of formatting errors. Simultaneously, performance data is stored fragmented across channels, hindering end-to-end attribution analysis. For example, it's impossible to track the complete user journey from clicking on a short video ad to a secondary search on a search engine and finally making a purchase, leading to low budget allocation accuracy and insufficient resource allocation to high-conversion channels. Furthermore, the lack of quantifiable evaluation models for the adaptability of creative elements relies heavily on the subjective judgment of optimization specialists. For instance, adjusting the tone and style of ad copy based solely on experience makes it difficult to develop standardized optimization strategies, resulting in performance fluctuations exceeding 20% ​​across different campaigns. Summary of the Invention

[0007] To address the aforementioned technical shortcomings, the purpose of this invention is to provide an advertising platform based on AI-driven dynamic creative optimization.

[0008] To solve the above-mentioned technical problems, the present invention adopts the following technical solution: The present invention provides an advertising platform based on AI dynamic creative optimization, including: a dynamic creative basic network construction module: used to deploy multimodal data collection and generation nodes on the target advertising platform, thereby constructing a full-scene dynamic creative intelligent generation network covering the target advertising platform.

[0009] Creative benchmark metric acquisition module: After the target advertising platform's full-scene dynamic creative intelligent generation network is completed, it acquires the effect data and adaptation data corresponding to each creative element in each advertisement at the current moment, and then analyzes and obtains the effect metrics and adaptation metrics corresponding to each creative element in each advertisement.

[0010] Creative Failure Assessment Module: This module is used to assess whether the creative elements in each advertisement have experienced a decline in effectiveness based on the corresponding performance metrics and compatibility metrics.

[0011] Alternative Creative Planning Module: When the effectiveness of a creative element in an advertisement diminishes, it analyzes the available alternative creative elements corresponding to that creative element in the advertisement, and then plans alternative solutions for each available alternative creative element corresponding to that creative element in the advertisement.

[0012] Creative Emergency Response Plan Execution Module: This module is used to execute the corresponding creative emergency response plan for an advertisement when all available alternative creative elements corresponding to the creative element in the advertisement experience a decline in effectiveness.

[0013] The beneficial effects of this invention are as follows: 1. Firstly, the embodiments of this invention achieve accurate prediction and real-time response to the decay of creative effects. By capturing the month-on-month changes in effect indicators and fluctuations in adaptation indicators in real time through multimodal data acquisition nodes, combined with a standardized evaluation model, an early warning can be triggered within 15% of the core indicators, allowing for intervention and optimization at least 6 hours earlier than traditional manual monitoring, and reducing the cost of invalid exposures by more than 40%. Simultaneously, the dynamic creative topology map can correlate with changes in user scenarios, improving the timeliness of creative adaptation adjustments by 60%, and maintaining an average click-through rate of over 90% of the baseline value.

[0014] 2. The embodiments of this invention firstly optimize the allocation of creative resources and the efficiency of alternative solutions. Based on a three-tiered allocation strategy using resource occupancy coefficients, resources can be dynamically allocated according to advertising value, increasing the resource utilization rate of core elements by 50% and avoiding resource waste for elements with low conversion potential. The alternative creative planning module uses attenuation risk value quantification screening, combined with real-time user preference data, to increase the matching degree between alternative solutions and the target audience to over 80%, resulting in an average 30% increase in conversion rate after switching, far exceeding the effect of traditional static backup libraries.

[0015] 3. The embodiments of this invention firstly and finally enhance emergency support and cross-channel collaboration capabilities in extreme scenarios. When all alternative creatives decay, the platform-level emergency engine can generate a completely new creative combination within 10 minutes, reducing the response cycle from 24 hours to within 1 hour, and reducing the loss of campaign interruptions during critical periods such as major promotions by 90%. The unified creative interface and data link across all channels realize automatic format adaptation and full-link attribution, improve the accuracy of budget allocation by 35%, control the fluctuation range of performance across different channels within 10%, and improve the overall ROI by 25%-40%. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a schematic diagram of the system module connections of the present invention. Detailed Implementation

[0018] 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.

[0019] Examples of embodiments of the present invention Figure 1 As shown, the AI-based dynamic creative optimization advertising platform includes: a dynamic creative basic network construction module, a creative benchmark indicator acquisition module, a creative failure assessment module, an alternative creative planning module, and a creative emergency response plan execution module.

[0020] The creative benchmark indicator acquisition module is connected to the dynamic creative basic network construction module and the creative failure assessment module, respectively, and the creative planning module is connected to the creative failure assessment module and the creative emergency response plan execution module, respectively.

[0021] Dynamic Creative Infrastructure Network Building Module: Used to deploy multimodal data collection and generation nodes on the target advertising platform, thereby building a dynamic creative intelligent generation network covering the entire target advertising platform.

[0022] In a specific embodiment, the construction of a full-scenario dynamic creative intelligent generation network covering the target advertising platform is specifically deployed as follows: Multimodal data collection nodes, including user behavior trackers, scene sensors, content recognizers, and generation nodes, are deployed on the mainstream channels covered by the target advertising platform; each generation node is assigned a unique creative identifier and establishes real-time communication with the platform's core management system through an encryption protocol to realize the distribution of creative parameters and the feedback of effect data; at the same time, based on the target audience characteristics, placement channel attributes, and creative dissemination paths of each advertisement, a dynamic creative topology map is constructed in the management system to form an intelligent generation network covering the entire scenario, and creative resource allocation strategies for each advertisement are formulated simultaneously.

[0023] In a specific embodiment, the process of formulating the creative resource allocation strategy for each advertisement is as follows: The resource occupancy coefficient of each advertisement on the target advertising platform is obtained, and this coefficient is compared with a preset three-tier allocation threshold. If the resource occupancy coefficient of an advertisement on the target advertising platform is less than or equal to the first-tier threshold, a "core element priority allocation" strategy is adopted. If the resource occupancy coefficient of an advertisement on the target advertising platform is between the first-tier and second-tier thresholds, a "balanced allocation" strategy is adopted. If the resource occupancy coefficient of an advertisement on the target advertising platform is greater than the second-tier threshold, a "flexible allocation" strategy is adopted.

[0024] It should be noted that the "core element priority allocation" strategy allocates 80% of creative resources to the core elements of the advertisement, such as the main product image and core copy, while focusing on 20% of resources for optimizing auxiliary elements to ensure maximum core appeal; the "balanced allocation" strategy allocates core and auxiliary element resources in a 6:4 ratio to enhance overall creative coordination while strengthening core competitiveness; the "flexible allocation" strategy does not set a fixed ratio and is dynamically adjusted based on real-time performance data, adding resources to elements that significantly improve click-through rate and conversion rate, and reducing investment in elements with declining performance to achieve precise resource matching.

[0025] It should be noted that the resource utilization coefficient = creative generation cost × target audience ÷ expected conversion value.

[0026] It should also be noted that the creative generation cost is calculated by summarizing the entire process of the ad creative from element design to multimodal content generation, including manual design time costs (calculated as designer's hourly wage × time invested), AI generation computing power costs (calculated as generation node runtime × unit computing power cost), and material copyright costs, etc., to obtain the total amount; the target audience is estimated by using the user profile system of the target advertising platform, combined with the user scale of the advertising channel, targeting conditions such as age, region, interest tags, and historical advertising reach data; the expected conversion value is based on the historical average order value, conversion rate, and user lifetime value of the product corresponding to the ad, and the expected total conversion revenue of a single ad is predicted by regression model. Finally, the above three values ​​are calculated into the resource occupancy coefficient using the formula "creative generation cost × target audience ÷ expected conversion value".

[0027] Creative benchmark metric acquisition module: After the target advertising platform's full-scene dynamic creative intelligent generation network is completed, it acquires the effect data and adaptation data corresponding to each creative element in each advertisement at the current moment, and then analyzes and obtains the effect metrics and adaptation metrics corresponding to each creative element in each advertisement.

[0028] In one specific embodiment, the performance data includes the click-through rate, average dwell time, secondary dissemination rate, and cost conversion rate of each creative element; the adaptation data includes the relevance of the element to the advertising theme, the fit with the placement scenario, the matching degree with user preferences, and the compatibility with platform rules.

[0029] It should be noted that performance data acquisition relies on the target advertising platform's real-time data interface and user behavior tracking system. Click-through rate (CTR) is calculated by the ratio of the number of times a creative element is clicked to the number of impressions. For example, if an image element is clicked 150 times out of 1000 impressions, the CTR is 15%. Average dwell time is calculated by recording the cumulative time users spend viewing ads containing that creative element and dividing that time by the number of valid views. Secondary dissemination rate is calculated by the ratio of the number of times users forward or share ads containing that element to the number of clicks. Cost-per-click (CPC) is calculated by dividing the total conversion value generated by the ad corresponding to that creative element by its ad spend, and this is synchronized to the platform's data dashboard in real time.

[0030] The acquisition of adaptation data requires a combination of AI analysis and platform rule matching. The relevance of elements to the ad theme is calculated by extracting the core semantics of the creative element's text / image using an NLP topic model and matching it with the pre-set ad theme keywords; a match rate of ≥80% indicates high relevance. The fit with the target context is determined by identifying the user's current scenario using scene sensors and comparing the element's historical performance data in the same scenario to generate a fit score of 0-100. The fit with user preferences is based on user profile tags, calculating the overlap between element features and user tags. Compatibility with platform rules is checked using a built-in platform rule verification engine to ensure elements meet requirements such as size and content compliance; the percentage of compliant items represents the compatibility score. All adaptation data is updated every 30 minutes to ensure timeliness.

[0031] In a specific embodiment, the analysis obtains the effect index and adaptation index corresponding to each creative element in each advertisement. The specific analysis process is as follows: the effect index and adaptation index corresponding to each creative element in each advertisement are standardized, and then substituted into the effect index calculation model and adaptation index calculation model respectively to obtain the effect index and adaptation index corresponding to each creative element in each advertisement.

[0032] It should be noted that the performance metrics = (click-through rate × 40% + average dwell time × 20% + secondary dissemination rate × 20% + cost conversion rate × 20%); the adaptation metrics = (theme relevance × 30% + scenario fit × 25% + user preference matching × 25% + platform compatibility × 20%).

[0033] Creative Failure Assessment Module: This module is used to assess whether the creative elements in each advertisement have experienced a decline in effectiveness based on the corresponding performance metrics and compatibility metrics.

[0034] In a specific embodiment, the evaluation process for assessing whether the creative elements in each advertisement experience effect decay is as follows: A1. Obtain the month-on-month change rate corresponding to the effect index of each creative element in each advertisement, and the fluctuation range corresponding to the adaptation index of each creative element in each advertisement, and compare the month-on-month change rate corresponding to the effect index of each creative element in each advertisement and the fluctuation range corresponding to the adaptation index with the set standard month-on-month change rate and standard fluctuation range, respectively.

[0035] It should be noted that the month-on-month change rate = (current period value performance index - previous period value performance index) ÷ previous period value performance index × 100%. The fluctuation range of the matching index = (maximum value matching index - minimum value matching index) ÷ average value matching index × 100%.

[0036] A2. If the month-on-month change rate of a certain creative element's performance indicator in an advertisement is greater than or equal to the set standard month-on-month change rate, and the fluctuation range of the adaptation indicator of the creative element in the advertisement is greater than or equal to the set standard fluctuation range, then it is determined that the creative element in the advertisement has experienced performance decay; otherwise, it is determined that the creative element in the advertisement has not experienced performance decay.

[0037] Alternative Creative Planning Module: When the effectiveness of a creative element in an advertisement diminishes, it analyzes the available alternative creative elements corresponding to that creative element in the advertisement, and then plans alternative solutions for each available alternative creative element corresponding to that creative element in the advertisement.

[0038] In a specific embodiment, the analysis of available alternative creative elements corresponding to the creative element in the advertisement is carried out as follows: Equivalent alternative elements corresponding to the creative element in the advertisement are selected from the platform's creative library to obtain the attenuation risk value corresponding to each equivalent alternative element. The attenuation risk value is calculated as: historical attenuation period × current adaptation index fluctuation range. The attenuation risk value corresponding to each equivalent alternative element is compared with a preset safety threshold. If the attenuation risk value corresponding to a certain equivalent alternative element is less than the preset safety threshold, then that equivalent alternative element is considered as a usable alternative creative element corresponding to the creative element in the advertisement.

[0039] In a specific embodiment, the planning of alternative creative elements corresponding to the creative element in the advertisement involves the following analysis process: sorting the attenuation risk values ​​of each equivalent alternative element from high to low, and selecting the top 3 equivalent alternative elements corresponding to the attenuation risk values ​​as candidate equivalent alternative elements; at the same time, executing parameters for the 3 candidate equivalent alternative elements, including switching trigger time, platform synchronization instructions, and effect monitoring nodes, to form a complete set of alternative solutions.

[0040] Creative Emergency Response Plan Execution Module: This module is used to execute the corresponding creative emergency response plan for an advertisement when all available alternative creative elements corresponding to the creative element in the advertisement experience a decline in effectiveness.

[0041] In a specific embodiment, the execution process of the creative emergency response plan corresponding to the advertisement is as follows: B1. If the attenuation risk value corresponding to each equivalent substitute element is compared with the preset safety threshold, and the attenuation risk value corresponding to each equivalent substitute element is less than the preset safety threshold, it indicates that the effect attenuation of each available substitute creative element corresponding to the creative element in the advertisement has occurred, and the third step of the creative emergency response plan corresponding to the advertisement is executed.

[0042] B2. First step: Activate the attenuation element isolation mechanism, send temporary blocking instructions to all distribution channels, and suspend the exposure of the creative element.

[0043] B3. The second step is to activate the platform-level emergency creative generation engine, which generates 5 new creative combinations within 10 minutes based on the core advertising appeal and real-time user data.

[0044] B4. The third step is to quickly test the newly generated creative combinations and select the options with a click-through rate greater than or equal to 120% of the benchmark value as the emergency main options.

[0045] B5. Fourth step: Issue a synchronized switching instruction to all channels, requiring creative updates to be completed within 30 minutes, and start 24-hour real-time monitoring, generating an hourly performance review report.

[0046] In a specific embodiment, the selection of schemes with a click-through rate (CTR) greater than or equal to a baseline value of 120% as the emergency primary scheme is carried out as follows: obtain the historical average CTR of the advertisement and set it as the baseline value; calculate the ratio of the test CTR of each emergency creative combination to the baseline value; if the ratio is ≥120%, it is included in the candidate pool; select the scheme with the highest conversion rate from the candidate pool as the emergency primary scheme; if the candidate pool is empty, the scheme with the highest matching degree in the platform's default backup creative library is activated.

[0047] It should be noted that the cumulative number of impressions and clicks for the ad within the period is retrieved through the data backend of the target advertising platform. If the ad contains multiple creative versions, the impressions of each version must be weighted and calculated, i.e., (Version A impressions × Version A click-through rate + Version B impressions × Version B click-through rate + ...) ÷ Total impressions. After removing abnormal data, such as malicious clicks from a single IP address or invalid impressions with an exposure duration of less than 0.5 seconds, the total number of clicks is divided by the total number of valid impressions to obtain the historical average click-through rate. The result is rounded to two decimal places and synchronized to the platform's baseline database.

[0048] The above description is merely an example and illustration of the concept of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described or use similar methods to replace them, as long as they do not deviate from the concept of the invention or exceed the scope defined in this specification, they should all fall within the protection scope of the present invention.

Claims

1. An advertising platform based on AI-driven dynamic creative optimization, characterized in that: include: Dynamic Creative Infrastructure Network Building Module: Used to deploy multimodal data collection and generation nodes on the target advertising platform, thereby building a dynamic creative intelligent generation network covering the entire target advertising platform. Creative benchmark metric acquisition module: After the full-scene dynamic creative intelligent generation network of the target advertising platform is completed, it is used to obtain the effect data and adaptation data corresponding to each creative element in each advertisement at the current moment, and then analyze and obtain the effect metric and adaptation metric corresponding to each creative element in each advertisement. Creative Failure Assessment Module: This module is used to assess whether the creative elements in each advertisement have experienced a decline in effectiveness based on the performance metrics and compatibility metrics corresponding to each creative element in each advertisement. Alternative Creative Planning Module: When the effect of a creative element in an advertisement diminishes, it analyzes the available alternative creative elements corresponding to that creative element in the advertisement, and then plans alternative solutions for each available alternative creative element corresponding to that creative element in the advertisement. Creative Emergency Response Plan Execution Module: This module is used to execute the corresponding creative emergency response plan for an advertisement when all available alternative creative elements corresponding to the creative element in the advertisement experience a decline in effectiveness.

2. The advertising platform based on AI dynamic creative optimization as described in claim 1, characterized in that, The specific deployment process for constructing a dynamic creative intelligent generation network covering the entire target advertising platform is as follows: Multimodal data collection nodes, including user behavior trackers, scene sensors, content recognizers, and generation nodes, are deployed on the mainstream channels covered by the target advertising platform. Each generation node is assigned a unique creative identifier and establishes real-time communication with the platform's core management system through an encrypted protocol to enable the distribution of creative parameters and the feedback of performance data. At the same time, based on the target audience characteristics, channel attributes, and creative dissemination paths of each advertisement, a dynamic creative topology map is constructed in the management system to form an intelligent generation network covering the entire scenario, and creative resource allocation strategies for each advertisement are formulated simultaneously.

3. The advertising platform based on AI dynamic creative optimization as described in claim 2, characterized in that, The specific process for formulating the creative resource allocation strategy for each advertisement is as follows: Obtain the resource utilization coefficient of each advertisement on the target advertising platform, and compare the resource utilization coefficient of each advertisement on the target advertising platform with the preset three-level allocation threshold: if the resource utilization coefficient of an advertisement on the target advertising platform is less than or equal to the first-level threshold, the "core element priority allocation" strategy is adopted. If the resource consumption coefficient of an advertisement on the target advertising platform is between the first-level threshold and the second-level threshold, then the "balanced allocation" strategy will be adopted. If the resource occupancy coefficient of an advertisement on the target advertising platform is greater than the secondary threshold, then the "elastic allocation" strategy will be adopted.

4. The advertising platform based on AI dynamic creative optimization as described in claim 3, characterized in that, The performance data includes the click-through rate, average dwell time, secondary dissemination rate, and cost-to-conversion rate of each creative element; the adaptation data includes the relevance of the element to the advertising theme, its fit with the advertising scenario, its match with user preferences, and its compatibility with platform rules.

5. The advertising platform based on AI dynamic creative optimization as described in claim 4, characterized in that, The analysis yielded the performance and suitability metrics for each creative element in each advertisement. The specific analysis process is as follows: The performance metrics and compatibility metrics corresponding to each creative element in each advertisement are standardized. Then, they are substituted into the performance metric calculation model and the compatibility metric calculation model respectively to obtain the performance metrics and compatibility metrics corresponding to each creative element in each advertisement.

6. The advertising platform based on AI dynamic creative optimization as described in claim 5, characterized in that, The evaluation process for determining whether the effectiveness of each creative element in each advertisement has diminished is as follows: A1. Obtain the month-on-month change rate of the performance index of each creative element in each advertisement, and the fluctuation range of the adaptation index of each creative element in each advertisement. Compare the month-on-month change rate of the performance index of each creative element in each advertisement and the fluctuation range of the adaptation index with the set standard month-on-month change rate and standard fluctuation range respectively. A2. If the month-on-month change rate of a certain creative element's performance indicator in an advertisement is greater than or equal to the set standard month-on-month change rate, and the fluctuation range of the adaptation indicator of the creative element in the advertisement is greater than or equal to the set standard fluctuation range, then it is determined that the creative element in the advertisement has experienced performance decay; otherwise, it is determined that the creative element in the advertisement has not experienced performance decay.

7. The advertising platform based on AI dynamic creative optimization as described in claim 6, characterized in that, The analysis of available alternative creative elements corresponding to the creative element in the advertisement is carried out in the following specific process: The platform's creative library is used to select equivalent alternative elements that correspond to the creative element in the advertisement, thereby obtaining the decay risk value corresponding to each equivalent alternative element. The decay risk value is calculated as historical decay period × current adaptation index fluctuation range. The decay risk value corresponding to each equivalent alternative element is compared with a preset safety threshold. If the decay risk value corresponding to an equivalent alternative element is less than the preset safety threshold, then the equivalent alternative element is used as a usable alternative creative element corresponding to the creative element in the advertisement.

8. The advertising platform based on AI dynamic creative optimization as described in claim 7, characterized in that, The specific analysis process for planning alternative creative elements corresponding to the creative element in the advertisement is as follows: The attenuation risk values ​​corresponding to each equivalent alternative element are sorted from high to low, and the equivalent alternative elements corresponding to the top 3 attenuation risk values ​​are selected as candidate equivalent alternative elements. At the same time, parameters are executed for the 3 candidate equivalent alternative elements, including switching trigger time, platform synchronization instructions and effect monitoring nodes, to form a complete set of alternative solutions.

9. The advertising platform based on AI dynamic creative optimization as described in claim 8, characterized in that, The specific execution process of the creative contingency response plan corresponding to the executed advertisement is as follows: B1. If the attenuation risk value corresponding to each equivalent substitute element is compared with the preset safety threshold, and the attenuation risk value corresponding to each equivalent substitute element is less than the preset safety threshold, it means that the effect attenuation of each available substitute creative element corresponding to the creative element in the advertisement has occurred. Then, the three steps of the creative emergency response plan corresponding to the advertisement are executed. B2. First step: Activate the attenuation element isolation mechanism, send temporary blocking instructions to all distribution channels, and suspend the exposure of the creative element; B3. The second step is to activate the platform-level emergency creative generation engine, which generates 5 new creative combinations within 10 minutes based on the core advertising appeal and real-time user data. B4. The third step is to quickly test the newly generated creative combinations and select the options with a click-through rate greater than or equal to 120% of the benchmark value as the emergency main options. B5. Fourth step: Issue a synchronized switching instruction to all channels, requiring creative updates to be completed within 30 minutes, and start 24-hour real-time monitoring, generating an hourly performance review report.

10. The advertising platform based on AI dynamic creative optimization as described in claim 9, characterized in that, The solutions with a click-through rate greater than or equal to 120% of the baseline value were selected as the primary emergency solutions. The specific selection process is as follows: Obtain the historical average click-through rate of the ad and set it as the baseline value. Calculate the ratio of the test click-through rate of each emergency creative combination to the baseline value. If the ratio is ≥120%, it is included in the candidate pool. Select the option with the highest conversion rate from the candidate pool as the emergency main option. If the candidate pool is empty, use the option with the highest matching degree in the platform's default backup creative library.

Images