Cross-platform advertisement putting strategy collaborative management method and system

Abstract

The invention discloses a cross-platform advertisement putting strategy collaborative management method and system, and relates to the technical field of advertisement putting, and the method comprises the steps: uploading a plurality of advertisement test samples to a putting platform test library for testing, obtaining real-time network information, and carrying out the evaluation of a test result, selecting the advertisement test sample with the highest advertisement putting return rate data as a main push advertisement, setting an advertisement putting strategy according to the hot time period of the platform, then pre-putting the main push advertisement on each platform according to the advertisement putting strategy to evaluate the pre-putting effect, establishing a revenue prediction model according to a deep reinforcement learning algorithm, and predicting the revenue of the main push advertisement. The method comprises the steps of predicting the income of advertising on each platform in real time, dynamically distributing a budget weight according to the predicted income data of the advertising on each platform to obtain an advertising strategy distribution table, and formulating an advertising income distribution contract according to a preset advertisement click rate and a conversion value. And according to the advertisement revenue distribution contract, calculating and distributing revenue to an account corresponding to the platform.

Description

Technical Field

[0001] The present invention relates to the field of advertising delivery technology, and in particular to a cross-platform advertising delivery strategy collaborative management method and system. Background Art

[0002] In recent years, with the optimization of the network environment and the popularization of mobile smart devices, short videos have rapidly developed as a new form of content delivery. With its creative and interesting features, it has gradually become the most popular form of self-media. Under such circumstances, many short video platforms have been born. With the increasing number of users on short video platforms, a large number of advertisers have been attracted to place video ads on short video platforms.

[0003] Existing technologies lack intelligent analysis of the relationship between advertising strategies and advertising revenue. Consequently, advertising decisions still rely primarily on intuition or instinct, or are based solely on platform traffic. This often makes it difficult to maximize the return on advertising investment and determine the optimal advertising strategy, especially when advertising is being delivered simultaneously on multiple platforms. Since each platform has different target audiences, the target customers for the ads are also inconsistent, requiring specific advertising strategies for each platform to maximize effectiveness.

[0004] In view of the above technical defects, a solution is now proposed. Summary of the Invention

[0005] The purpose of the present invention is to optimize advertising delivery strategies according to different platforms, and at the same time provide an automatic revenue settlement function, thereby improving the conversion rate of advertising delivery and avoiding disputes.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: a cross-platform advertising delivery strategy collaborative management method, comprising the following steps:

[0007] Step 1: Obtain multiple ad test samples for the same product and upload them to the test library of the delivery platform for testing. Obtain real-time network information to evaluate the test results and obtain advertising return rate data. The ad test sample with the highest advertising return rate data will be selected as the main promotion ad, and the advertising delivery strategy will be set according to the platform's popular time periods.

[0008] Step 2: Pre-launch the featured ad on each platform according to the ad placement strategy. Obtain network information from each platform, along with ad content. Extract multi-dimensional features based on the network information to construct user profiles and ad context features to evaluate the effectiveness of the pre-launch.

[0009] Step 3: Collect historical advertising revenue data and historical advertising delivery strategies for each platform in a time series, establish a revenue prediction model based on a deep reinforcement learning algorithm, and periodically update the model parameters based on real-time data to predict the revenue of advertising on each platform in real time;

[0010] Step 4: Obtain the current advertising strategy, dynamically allocate budget weights based on the revenue data predicted by each platform's advertisements, obtain an advertising strategy allocation table, and update the advertising strategy allocation table accordingly based on the real-time updated revenue data predicted by each platform's advertisements;

[0011] Step 5: Develop an advertising revenue distribution contract based on the preset ad click volume and conversion value, obtain advertising consumption data from each platform, and calculate and distribute revenue to the corresponding account of the platform according to the advertising revenue distribution contract.

[0012] Furthermore, the specific process of setting up an advertising delivery strategy is as follows:

[0013] S101. Upload multiple advertising test samples for the same product to the delivery platform test library, obtain the popular time periods and unpopular time periods of the delivery platform as the advertising delivery periods, and test delivery the multiple advertising test samples in the popular time periods and unpopular time periods according to the basic delivery strategy;

[0014] S102: Acquire advertising test delivery data for each advertising test sample, wherein the advertising test delivery data includes click volume, cost per thousand impressions, cost per click, and cost per conversion;

[0015] The CTR is the ratio of the number of times an ad is clicked to the number of times the ad is displayed, that is, CTR = number of clicks / number of displays;

[0016] The cost per thousand impressions (CPM) is the ratio of the advertising fee to the number of ad impressions, that is, CPM = advertising fee / number of impressions;

[0017] The cost per click (CPC) is the ratio of the advertising cost to the number of clicks on the ad, that is, CPC = advertising cost / number of clicks;

[0018] Cost per conversion (CPA) is the ratio of the advertising cost to the number of conversions of the advertising, that is, CPA = advertising cost / number of conversions;

[0019] S103, calculating the advertising return rate data Ki for each advertising test sample in different advertising delivery periods according to the following formula: Among them, e1, e2, e3 and e4 are preset proportional coefficients, d is a constant, and the specific value of d is set according to the popularity value of the advertising period. The advertising return rate data is used to feedback the comprehensive income of the advertising test sample during the test delivery process. The larger the advertising return rate data, the higher the comprehensive income of the advertising test sample. Conversely, the smaller the advertising return rate data, the lower the comprehensive income of the advertising test sample.

[0020] S104: Select the advertising test sample with the highest advertising return rate data as the main advertising, and set the advertising delivery strategy based on the revenue ratio of popular time periods and unpopular time periods.

[0021] Furthermore, the specific process of evaluating the pre-launch effect is as follows;

[0022] S201. Obtain network information from each network platform, including user behavior logs, including user personal information and user behavior. The user personal information includes user age, gender, occupation, and IP location. The user behavior includes user click locations, dwell time, and page jump paths.

[0023] S202: Analyze user activity time periods based on timestamp clustering in user behavior logs, identify user behavior paths, calculate funnel conversion rates based on conversion rates at each step from click to order placement, and identify churn nodes.

[0024] S203: Create a basic user profile based on user personal information, add tags to the user profile based on user behavior preferences and value stratification, obtain advertising content, and dynamically generate advertising matching tags based on the user's real-time behavior, specifically the search keywords and the length and number of times the product is browsed.

[0025] S204: Then, a logistic regression model is trained to predict the user conversion rate, and the pre-delivery effect is evaluated based on the user conversion rate. When the user conversion rate is greater than a preset conversion rate threshold, it indicates that the pre-delivery effect has achieved an ideal effect.

[0026] Furthermore, the specific process of establishing the revenue forecasting model is as follows:

[0027] S301. Obtain historical advertising revenue data and historical advertising delivery strategies for each platform to be delivered. Extract features from the historical advertising delivery strategies and assign feature labels. Organize the historical advertising revenue data according to time series to generate a revenue data fluctuation table. Align the feature labels with the revenue data fluctuation table one-to-one, integrate the data to generate training samples, and divide the generated training samples into a training set and a test set in a ratio of 8:2.

[0028] S302, download the weight file and load it onto the corresponding network to initialize the migration network parameters, and determine the number of hidden layer nodes of the BP neural network model based on the number of samples in the training set;

[0029] S303. Modify the last fully connected layer of the network, keep the input unchanged, set the output to the predicted profit data, initialize the weights of the last layer, use the gradient descent algorithm for learning, and use fixed step size decay to optimize the training parameters. Retrain the entire network to obtain the profit prediction model;

[0030] S304: During the training process, small batches of training samples are randomly drawn from the training set without duplication. Drawing all the training samples in the training set constitutes one training cycle. The training is completed after a certain number of iterations, and the performance of the return rate prediction model is evaluated using the test set.

[0031] S305: Substitute the current advertising delivery strategy into the traffic prediction model to obtain predicted revenue data.

[0032] Furthermore, the specific process of obtaining the advertisement delivery strategy allocation table is as follows:

[0033] S401. Obtain the current advertising strategy and the predicted revenue data for advertising on each platform. Determine the total advertising cycle and total advertising resources based on the business needs and marketing plan in the current advertising strategy. Divide the total advertising cycle into a preset period, a main promotion period, and a sprint period, and set optimization targets for each period.

[0034] S402. Assign preliminary weights based on the specific values ​​of the income data. The expression is:

[0035] Where Wj is the initial weight of the jth optimization target, Wj is a specific value set according to the specific value of the income data and is positively correlated with the income data, Ej is the information entropy value of the jth optimization target, λ is the adjustment coefficient for adjusting the difference in the weight distribution of the information entropy value of the optimization target, γ is the adjustment coefficient for controlling the degree of weakening of the weight of the high entropy value target, Emax is the maximum value of the information entropy of all optimization targets, and k is the preset proportional coefficient;

[0036] S403. Divide the advertising delivery cycle in the current advertising delivery strategy according to the preliminary weights and allocate resources to obtain the adaptive advertising delivery strategy of each platform. According to the platform popularity value, integrate and obtain an advertising delivery strategy allocation table.

[0037] Furthermore, in step 5, a consortium chain is established between the platforms, an advertising revenue distribution contract is deployed on the consortium chain, and each platform encrypts and uploads the advertising data to the chain for storage;

[0038] The advertising revenue distribution contract receives updated advertising data, automatically triggers calculations, and distributes revenue to designated accounts without the need for manual intervention.

[0039] The present invention also provides a cross-platform advertising delivery strategy collaborative management system, comprising a delivery strategy planning unit, a delivery effect evaluation unit, an advertising revenue prediction unit, a delivery strategy allocation unit, and a revenue calculation unit, wherein:

[0040] The delivery strategy planning unit is used to obtain multiple advertising test samples for the same product, upload them to the delivery platform test library for testing, obtain real-time network information to evaluate the test results, obtain advertising delivery return rate data, select the advertising test sample with the highest advertising delivery return rate data as the main promotion advertisement, and set the advertising delivery strategy according to the platform's popular time period;

[0041] The delivery effect evaluation unit is used to pre-deliver the main advertisement on various platforms according to the advertisement delivery strategy, obtain network information of each network platform, and simultaneously obtain the advertisement content. Based on the network information, multi-dimensional features are extracted to construct user profiles and advertisement context features to evaluate the pre-delivery effect;

[0042] The advertising revenue prediction unit is used to collect historical advertising revenue data and historical advertising delivery strategies of each platform in a time series, establish a revenue prediction model based on the deep reinforcement learning algorithm, and periodically update the model parameters based on real-time data to predict the revenue of advertising on each platform in real time;

[0043] The delivery strategy allocation unit is used to obtain the current advertising delivery strategy, dynamically allocate budget weights based on the revenue data predicted by advertisements on each platform, obtain an advertising delivery strategy allocation table, and update the advertising delivery strategy allocation table accordingly based on the real-time updated revenue data predicted by advertisements on each platform;

[0044] The revenue calculation unit is used to formulate an advertising revenue distribution contract based on the preset advertising click volume and conversion value, obtain the advertising consumption data of each platform, and calculate and distribute the revenue to the corresponding account of the platform according to the advertising revenue distribution contract.

[0045] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are:

[0046] The cross-platform advertising delivery strategy collaborative management method and system uploads multiple advertising test samples to the delivery platform test library for testing, obtains real-time network information to evaluate the test results, selects the advertising test sample with the highest advertising delivery return rate data as the main recommended advertisement, and sets the advertising delivery strategy according to the platform's popular time period. The main recommended advertisement is then pre-delivered on each platform according to the advertising delivery strategy to evaluate the pre-delivery effect. A revenue prediction model is established based on a deep reinforcement learning algorithm to predict the revenue of advertising on each platform in real time. Budget weights are dynamically allocated based on the revenue data predicted by the advertisements on each platform to obtain an advertising delivery strategy allocation table. At the same time, an advertising revenue distribution contract is formulated based on the preset advertising click volume and conversion value. The revenue is calculated and allocated to the corresponding account of the platform according to the advertising revenue distribution contract. This method can not only optimize the advertising delivery strategy according to different platforms, but also provide an automatic revenue settlement function, thereby improving the conversion rate of advertising delivery and avoiding disputes. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 Shown is a schematic diagram of the overall method flow of the present invention;

[0048] Figure 2 Shown is a schematic diagram of the overall system structure of the present invention. DETAILED DESCRIPTION

[0049] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0050] Example 1:

[0051] like Figure 1 As shown, a cross-platform advertising delivery strategy collaborative management method is characterized by comprising the following steps:

[0052] Step 1: Obtain multiple ad test samples for the same product and upload them to the test library of the delivery platform for testing. Obtain real-time network information to evaluate the test results and obtain advertising return rate data. The ad test sample with the highest advertising return rate data will be selected as the main promotion ad, and the advertising delivery strategy will be set according to the platform's popular time periods.

[0053] The specific process of setting up an ad delivery strategy is as follows:

[0054] S101. Upload multiple advertising test samples for the same product to the delivery platform test library, obtain the popular time periods and unpopular time periods of the delivery platform as the advertising delivery periods, and test delivery the multiple advertising test samples in the popular time periods and unpopular time periods according to the basic delivery strategy;

[0055] S102: Acquire advertising test delivery data for each advertising test sample, wherein the advertising test delivery data includes click volume, cost per thousand impressions, cost per click, and cost per conversion;

[0056] The CTR is the ratio of the number of times an ad is clicked to the number of times the ad is displayed, that is, CTR = number of clicks / number of displays;

[0057] The cost per thousand impressions (CPM) is the ratio of the advertising fee to the number of ad impressions, that is, CPM = advertising fee / number of impressions;

[0058] The cost per click (CPC) is the ratio of the advertising cost to the number of clicks on the ad, that is, CPC = advertising cost / number of clicks;

[0059] Cost per conversion (CPA) is the ratio of the advertising cost to the number of conversions of the advertising, that is, CPA = advertising cost / number of conversions;

[0060] S103, calculating the advertising return rate data Ki for each advertising test sample in different advertising delivery periods according to the following formula: Among them, e1, e2, e3 and e4 are preset proportional coefficients, d is a constant, and the specific value of d is set according to the popularity value of the advertising period. The advertising return rate data is used to feedback the comprehensive income of the advertising test sample during the test delivery process. The larger the advertising return rate data, the higher the comprehensive income of the advertising test sample. Conversely, the smaller the advertising return rate data, the lower the comprehensive income of the advertising test sample.

[0061] S104: Select the advertising test sample with the highest advertising return rate data as the main advertising, and set the advertising delivery strategy based on the revenue ratio of popular time periods and unpopular time periods.

[0062] Step 2: Pre-launch the featured ad on each platform according to the ad placement strategy. Obtain network information from each platform, along with ad content. Extract multi-dimensional features based on the network information to construct user profiles and ad context features to evaluate the effectiveness of the pre-launch.

[0063] The specific process of evaluating the pre-launch effect is as follows;

[0064] S201. Obtain network information from each network platform, including user behavior logs, including user personal information and user behavior. The user personal information includes user age, gender, occupation, and IP location. The user behavior includes user click locations, dwell time, and page jump paths.

[0065] S202: Analyze user activity time periods based on timestamp clustering in user behavior logs, identify user behavior paths, calculate funnel conversion rates based on conversion rates at each step from click to order placement, and identify churn nodes.

[0066] S203: Create a basic user profile based on user personal information, add tags to the user profile based on user behavior preferences and value stratification, obtain advertising content, and dynamically generate advertising matching tags based on the user's real-time behavior, specifically the search keywords and the length and number of times the product is browsed.

[0067] S204: Then, a logistic regression model is trained to predict the user conversion rate, and the pre-delivery effect is evaluated based on the user conversion rate. When the user conversion rate is greater than a preset conversion rate threshold, it indicates that the pre-delivery effect has achieved an ideal effect.

[0068] Step 3: Collect historical advertising revenue data and historical advertising delivery strategies for each platform in a time series, establish a revenue prediction model based on a deep reinforcement learning algorithm, and periodically update the model parameters based on real-time data to predict the revenue of advertising on each platform in real time;

[0069] The specific process of establishing a revenue forecasting model is as follows:

[0070] S301. Obtain historical advertising revenue data and historical advertising delivery strategies for each platform to be delivered. Extract features from the historical advertising delivery strategies and assign feature labels. Organize the historical advertising revenue data according to time series to generate a revenue data fluctuation table. Align the feature labels with the revenue data fluctuation table one-to-one, integrate the data to generate training samples, and divide the generated training samples into a training set and a test set in a ratio of 8:2.

[0071] S302, download the weight file and load it onto the corresponding network to initialize the migration network parameters, and determine the number of hidden layer nodes of the BP neural network model based on the number of samples in the training set;

[0072] S303. Modify the last fully connected layer of the network, keep the input unchanged, set the output to the predicted profit data, initialize the weights of the last layer, use the gradient descent algorithm for learning, and use fixed step size decay to optimize the training parameters. Retrain the entire network to obtain the profit prediction model;

[0073] S304: During the training process, small batches of training samples are randomly drawn from the training set without duplication. Drawing all the training samples in the training set constitutes one training cycle. The training is completed after a certain number of iterations, and the performance of the return rate prediction model is evaluated using the test set.

[0074] S305: Substitute the current advertising delivery strategy into the traffic prediction model to obtain predicted revenue data.

[0075] Step 4: Obtain the current advertising strategy, dynamically allocate budget weights based on the revenue data predicted by each platform's advertisements, obtain an advertising strategy allocation table, and update the advertising strategy allocation table accordingly based on the real-time updated revenue data predicted by each platform's advertisements;

[0076] The specific process of obtaining the advertising delivery strategy allocation table is as follows:

[0077] S401. Obtain the current advertising strategy and the predicted revenue data for advertising on each platform. Determine the total advertising cycle and total advertising resources based on the business needs and marketing plan in the current advertising strategy. Divide the total advertising cycle into a preset period, a main promotion period, and a sprint period, and set optimization targets for each period.

[0078] S402. Assign preliminary weights based on the specific values ​​of the income data. The expression is:

[0079] Where Wj is the initial weight of the jth optimization target, Wj is a specific value set according to the specific value of the income data and is positively correlated with the income data, Ej is the information entropy value of the jth optimization target, λ is the adjustment coefficient for adjusting the difference in the weight distribution of the information entropy value of the optimization target, γ is the adjustment coefficient for controlling the degree of weakening of the weight of the high entropy value target, Emax is the maximum value of the information entropy of all optimization targets, and k is the preset proportional coefficient;

[0080] S403. Divide the advertising delivery cycle in the current advertising delivery strategy according to the preliminary weights and allocate resources to obtain the adaptive advertising delivery strategy of each platform. According to the platform popularity value, integrate and obtain an advertising delivery strategy allocation table.

[0081] Step 5: Develop an advertising revenue distribution contract based on the preset ad click volume and conversion value, obtain advertising consumption data from each platform, and calculate and distribute revenue to the corresponding account of the platform according to the advertising revenue distribution contract.

[0082] In step 5, a consortium chain is established between the platforms, an advertising revenue distribution contract is deployed on the consortium chain, and each platform encrypts and uploads the advertising data to the chain for storage;

[0083] The advertising revenue distribution contract receives updated advertising data, automatically triggers calculations, and distributes revenue to designated accounts without the need for manual intervention.

[0084] The present invention uploads multiple advertising test samples to the test library of the delivery platform for testing, obtains real-time network information to evaluate the test results, selects the advertising test sample with the highest advertising delivery return rate data as the main promotion advertisement, and sets the advertising delivery strategy according to the popular time period of the platform. The main promotion advertisement is then pre-delivered on each platform according to the advertising delivery strategy to evaluate the pre-delivery effect. A revenue prediction model is established based on the deep reinforcement learning algorithm to predict the revenue of advertising on each platform in real time. Budget weights are dynamically allocated according to the revenue data predicted by the advertisements on each platform to obtain an advertising delivery strategy allocation table. At the same time, an advertising revenue distribution contract is formulated according to the preset advertising click volume and conversion value, and the revenue is calculated and distributed to the corresponding account of the platform according to the advertising revenue distribution contract.

[0085] Example 2:

[0086] like Figure 2 As shown, a cross-platform advertising delivery strategy collaborative management system is characterized by including a delivery strategy planning unit, a delivery effect evaluation unit, an advertising revenue prediction unit, a delivery strategy allocation unit and a revenue calculation unit, wherein:

[0087] The delivery strategy planning unit is used to obtain multiple advertising test samples for the same product, upload them to the delivery platform test library for testing, obtain real-time network information to evaluate the test results, obtain advertising delivery return rate data, select the advertising test sample with the highest advertising delivery return rate data as the main promotion advertisement, and set the advertising delivery strategy according to the platform's popular time period;

[0088] The delivery effect evaluation unit is used to pre-deliver the main advertisement on various platforms according to the advertisement delivery strategy, obtain network information of each network platform, and simultaneously obtain the advertisement content. Based on the network information, multi-dimensional features are extracted to construct user profiles and advertisement context features to evaluate the pre-delivery effect;

[0089] The advertising revenue prediction unit is used to collect historical advertising revenue data and historical advertising delivery strategies of each platform in a time series, establish a revenue prediction model based on the deep reinforcement learning algorithm, and periodically update the model parameters based on real-time data to predict the revenue of advertising on each platform in real time;

[0090] The delivery strategy allocation unit is used to obtain the current advertising delivery strategy, dynamically allocate budget weights based on the revenue data predicted by advertisements on each platform, obtain an advertising delivery strategy allocation table, and update the advertising delivery strategy allocation table accordingly based on the real-time updated revenue data predicted by advertisements on each platform;

[0091] The revenue calculation unit is used to formulate an advertising revenue distribution contract based on the preset advertising click volume and conversion value, obtain the advertising consumption data of each platform, and calculate and distribute the revenue to the corresponding account of the platform according to the advertising revenue distribution contract.

[0092] The threshold is set to facilitate comparison. The size of the threshold depends on the amount of sample data and the number of bases set by technicians in this field for each set of sample data; as long as it does not affect the proportional relationship between the parameter and the quantized value.

[0093] The above formulas are all dimensionless and numerical calculations. The formulas are obtained by collecting a large amount of data and performing software simulation to obtain the most recent real situation. The preset parameters in the formulas are set by those skilled in the art according to actual conditions.

[0094] In the two embodiments provided in this application, it should be understood that the disclosed methods and systems can be implemented in other ways. For example, the division of modules is merely a logical function division, and in actual implementation, there may be other division methods, such as multiple modules or components can be combined or integrated into another system, or some features can be ignored or not performed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed may be through some interfaces, and the indirect coupling or communication connection of modules may be electrical or other forms.

[0095] The above description is only a preferred specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with the technical field, within the technical scope disclosed by the present invention, who makes equivalent replacements or changes based on the technical solution and inventive concept of the present invention, should be covered by the scope of protection of the present invention.

Claims

1. A cross-platform advertising delivery strategy collaborative management method, characterized in that: The following steps are involved: Step 1: Obtain multiple ad test samples for the same product and upload them to the test library of the delivery platform for testing. Obtain real-time network information to evaluate the test results and obtain advertising return rate data. The ad test sample with the highest advertising return rate data will be selected as the main promotion ad, and the advertising delivery strategy will be set according to the platform's popular time periods. Step 2: Pre-launch the featured ad on each platform according to the ad placement strategy. Obtain network information from each platform, along with ad content. Extract multi-dimensional features based on the network information to construct user profiles and ad context features to evaluate the effectiveness of the pre-launch. Step 3: Collect historical advertising revenue data and historical advertising delivery strategies for each platform in a time series, establish a revenue prediction model based on a deep reinforcement learning algorithm, and periodically update the model parameters based on real-time data to predict the revenue of advertising on each platform in real time; Step 4: Obtain the current advertising strategy, dynamically allocate budget weights based on the revenue data predicted by each platform's advertisements, obtain an advertising strategy allocation table, and update the advertising strategy allocation table accordingly based on the real-time updated revenue data predicted by each platform's advertisements; Step 5: Develop an advertising revenue distribution contract based on the preset ad click volume and conversion value, obtain advertising consumption data from each platform, and calculate and distribute revenue to the corresponding account of the platform according to the advertising revenue distribution contract.

2. A cross-platform advertising delivery strategy collaborative management method according to claim 1, characterized in that: The specific process of setting up an ad delivery strategy is as follows: S101. Upload multiple advertising test samples for the same product to the delivery platform test library, obtain the popular time periods and unpopular time periods of the delivery platform as the advertising delivery periods, and test delivery the multiple advertising test samples in the popular time periods and unpopular time periods according to the basic delivery strategy; S102: Acquire advertising test delivery data for each advertising test sample, wherein the advertising test delivery data includes click volume, cost per thousand impressions, cost per click, and cost per conversion; The CTR is the ratio of the number of times an ad is clicked to the number of times the ad is displayed, that is, CTR = number of clicks / number of displays; The cost per thousand impressions (CPM) is the ratio of the advertising fee to the number of ad impressions, that is, CPM = advertising fee / number of impressions; The cost per click (CPC) is the ratio of the advertising cost to the number of clicks on the ad, that is, CPC = advertising cost / number of clicks; Cost per conversion (CPA) is the ratio of the advertising cost to the number of conversions of the advertising, that is, CPA = advertising cost / number of conversions; S103, calculating the advertising return rate data Ki for each advertising test sample in different advertising delivery periods according to the following formula: Among them, e1, e2, e3 and e4 are preset proportional coefficients, d is a constant, and the specific value of d is set according to the popularity value of the advertising delivery period. The advertising delivery return rate data is used to feedback the comprehensive benefits of the advertising test sample during the test delivery process; S104: Select the advertising test sample with the highest advertising return rate data as the main advertising, and set the advertising delivery strategy based on the revenue ratio of popular time periods and unpopular time periods.

3. A cross-platform advertising delivery strategy collaborative management method according to claim 1, characterized in that: The specific process of evaluating the pre-launch effect is as follows; S201. Obtain network information from each network platform, including user behavior logs, including user personal information and user behavior. The user personal information includes user age, gender, occupation, and IP location. The user behavior includes user click locations, dwell time, and page jump paths. S202: Analyze user activity time periods based on user behavior logs through timestamp clustering, identify user behavior paths, calculate funnel conversion rates based on conversion rates at each step from click to order placement, and identify churn nodes. S203: Create a basic user profile based on user personal information, add tags to the user profile based on user behavior preferences and value stratification, obtain advertising content, and dynamically generate advertising matching tags based on the user's real-time behavior, specifically the search keywords and the length and number of times the product is browsed. S204: Then, a logistic regression model is trained to predict the user conversion rate, and the pre-delivery effect is evaluated based on the user conversion rate. When the user conversion rate is greater than a preset conversion rate threshold, it indicates that the pre-delivery effect has achieved an ideal effect.

4. A cross-platform advertising delivery strategy collaborative management method according to claim 1, characterized in that: The specific process of establishing a revenue forecasting model is as follows: S301. Obtain historical advertising revenue data and historical advertising delivery strategies for each platform to be delivered. Extract features from the historical advertising delivery strategies and assign feature labels. Organize the historical advertising revenue data according to time series to generate a revenue data fluctuation table. Align the feature labels with the revenue data fluctuation table one-to-one, integrate the data to generate training samples, and divide the generated training samples into a training set and a test set in a ratio of 8:

2. S302, download the weight file and load it onto the corresponding network to initialize the migration network parameters, and determine the number of hidden layer nodes of the BP neural network model based on the number of samples in the training set; S303. Modify the last fully connected layer of the network, keep the input unchanged, set the output to the predicted profit data, initialize the weights of the last layer, use the gradient descent algorithm for learning, and use fixed step size decay to optimize the training parameters. Retrain the entire network to obtain the profit prediction model; S304: During the training process, small batches of training samples are randomly drawn from the training set without duplication. Drawing all the training samples in the training set constitutes one training cycle. The training is completed after a certain number of iterations, and the performance of the return rate prediction model is evaluated using the test set. S305: Substitute the current advertising delivery strategy into the traffic prediction model to obtain predicted revenue data.

5. A cross-platform advertising delivery strategy collaborative management method according to claim 1, characterized in that: The specific process of obtaining the advertising delivery strategy allocation table is as follows: S401. Obtain the current advertising strategy and the predicted revenue data for advertising on each platform. Determine the total advertising cycle and total advertising resources based on the business needs and marketing plan in the current advertising strategy. Divide the total advertising cycle into a preset period, a main promotion period, and a sprint period, and set optimization targets for each period. S402. Assign preliminary weights based on the specific values ​​of the income data. The expression is: Where Wj is the initial weight of the jth optimization target, Wj is a specific value set according to the specific value of the income data and is positively correlated with the income data, Ej is the information entropy value of the jth optimization target, λ is the adjustment coefficient for adjusting the difference in the weight distribution of the information entropy value of the optimization target, γ is the adjustment coefficient for controlling the degree of weakening of the weight of the high entropy value target, Emax is the maximum value of the information entropy of all optimization targets, and k is the preset proportional coefficient; S403. Divide the advertising delivery cycle in the current advertising delivery strategy according to the preliminary weights and allocate resources to obtain the adaptive advertising delivery strategy of each platform. According to the platform popularity value, integrate and obtain an advertising delivery strategy allocation table.

6. A cross-platform advertising delivery strategy collaborative management method according to claim 1, characterized in that: In step 5, a consortium chain is established between the platforms, an advertising revenue distribution contract is deployed on the consortium chain, and each platform encrypts and uploads the advertising data to the chain for storage; The advertising revenue distribution contract receives updated advertising data, automatically triggers calculations, and distributes revenue to designated accounts without the need for manual intervention.

7. A cross-platform advertising delivery strategy collaborative management system, characterized in that: It includes a delivery strategy planning unit, a delivery effect evaluation unit, an advertising revenue forecasting unit, a delivery strategy allocation unit, and a revenue calculation unit, among which: The delivery strategy planning unit is used to obtain multiple advertising test samples for the same product, upload them to the delivery platform test library for testing, obtain real-time network information to evaluate the test results, obtain advertising delivery return rate data, select the advertising test sample with the highest advertising delivery return rate data as the main promotion advertisement, and set the advertising delivery strategy according to the platform's popular time period; The delivery effect evaluation unit is used to pre-deliver the main advertisement on various platforms according to the advertisement delivery strategy, obtain network information of each network platform, and simultaneously obtain the advertisement content. Based on the network information, multi-dimensional features are extracted to construct user profiles and advertisement context features to evaluate the pre-delivery effect; The advertising revenue prediction unit is used to collect historical advertising revenue data and historical advertising delivery strategies of each platform in a time series, establish a revenue prediction model based on the deep reinforcement learning algorithm, and periodically update the model parameters based on real-time data to predict the revenue of advertising on each platform in real time; The delivery strategy allocation unit is used to obtain the current advertising delivery strategy, dynamically allocate budget weights based on the revenue data predicted by advertisements on each platform, obtain an advertising delivery strategy allocation table, and update the advertising delivery strategy allocation table accordingly based on the real-time updated revenue data predicted by advertisements on each platform; The revenue calculation unit is used to formulate an advertising revenue distribution contract based on the preset advertising click volume and conversion value, obtain the advertising consumption data of each platform, and calculate and distribute the revenue to the corresponding account of the platform according to the advertising revenue distribution contract.

Images