Advertisement traffic delivery method and device and storage medium

Abstract

The invention provides an advertisement flow putting method and device and a storage medium, and relates to the field of advertisement flow putting, and the method comprises the following steps: S100, data collection: obtaining a historical price, a click rate, an exposure rate and a conversion rate of a corresponding advertisement position from an advertisement server according to the type of an advertisement needing to be put; according to the method, the multiple linear regression model is adopted, various factors such as the historical price, the click rate, the exposure rate and the conversion rate of the advertisement space and the user equipment performance are comprehensively considered, a more accurate traffic prediction model is constructed, the labor intensity and risks of manual calculation of people are reduced, traffic can be predicted, and the traffic prediction efficiency is improved. In addition, the advertisement putting strategy can be adjusted through the traffic expected value, the traffic expected value is predicted by knowing the equipment performance information of the user, the possibility that the user loses and the conversion rate is reduced due to the fact that the user loads slow or lagged advertisements due to the equipment problem is reduced, the advertisement putting strategy is optimized, and the user experience is improved. And the user experience and the conversion rate are improved.

Description

Technical Field

[0001] The present invention relates to the field of advertising traffic delivery, and in particular to an advertising traffic delivery method, device and storage medium. Background Art

[0002] In traditional advertising traffic delivery plans, ad slot operators manually calculate the required traffic based on the demand for advertising and the estimated daily traffic of the ad slots, and formulate an advertising traffic delivery plan for the ad slots to be delivered. Usually, ad slot operators manually calculate based on their own experience or fixed formulas, which is a heavy workload and requires calculation every time. Not only is the workload heavy, but there is also a high risk of delivering advertisements with large traffic to multiple ad slots at one time. If the publicity effect is not good, the business value will be greatly lost.

[0003] Therefore, we have made improvements to this and proposed an advertising traffic delivery method, device and storage medium. Summary of the invention

[0004] The purpose of the present invention is to address the problems of the current background technology.

[0005] In order to achieve the above-mentioned purpose of the invention, the present invention provides the following advertising traffic delivery method, device and storage medium to improve the above-mentioned problem.

[0006] The specific application is as follows: A method for delivering advertising traffic includes the following steps: S100, data collection: according to the type of advertisements to be placed, the historical price, click-through rate, exposure rate and conversion rate of the corresponding advertisement slots are obtained from the advertisement server. The hardware data of the target media platform users are directly collected from the application through SDK integration. The hardware data includes screen resolution, processor frequency, running memory, network speed and energy consumption data. S200, calculating the advertising traffic prediction value: using the data collected in step S100 to build an advertising traffic prediction model to predict the advertising traffic; S200 specifically includes the following steps: S201. Construct prediction function: XGBoost is used as a prediction model, and its basic form can be expressed as ,in is the predicted advertising traffic, is the input feature vector, It is the nonlinear mapping function provided by the trained XGBoost model; S202, application model: taking new ad space data and user device hardware data as input features , use the trained XGBoost model to predict the advertising traffic ; S300, calculating a comprehensive index of hardware data: calculating a comprehensive index based on the hardware data of the user's device to evaluate the display effect of the advertisement on different devices; S400, traffic test: conduct a small-scale test on the selected ad slot in a real environment according to the first preset traffic, set the initial test ratio to 5% to 10% of the total user group, create an A / B test framework, and respectively include a control group and an experimental group, and obtain the test results; S500, using a traffic value model to calculate the expected traffic value: calculating the test results to obtain the expected traffic value of each advertising position; S500 specifically includes the following steps: S501, model structure: Expected value of traffic formula: ,in, is the click-through rate, is the conversion rate, It is a comprehensive index of the user's device hardware data. It's the exposure rate. is the weight coefficient, is the random error term; S600, Traffic Ratio and Threshold Comparison and Strategy Optimization: Calculating Expected Traffic Value With target value threshold The traffic ratio is calculated by adjusting the advertising traffic according to the traffic ratio. The calculation formula is: ; S700. Optimize advertising delivery effects: Introduce a comprehensive index. When it is detected that the user device of a certain ad slot is a low-quality device, the algorithm model is adjusted to output the optimal ad format settings.

[0007] As a preferred technical solution of the present application, S100 specifically includes the following steps: S101. Data preprocessing: Data cleaning: remove outliers, fill in missing values, and correct erroneous records; Standardization and normalization: Standardize and normalize data of different dimensions so that they are in the same numerical range for subsequent modeling.

[0008] As a preferred technical solution of this application, in S201, the goal of XGBoost is to minimize the following loss function: ,in is the loss function of a single sample, measuring the The true labels of samples With the predicted value The difference between is a regularization term used to prevent overfitting.

[0009] As a preferred technical solution of the present application, S300 specifically includes the following steps: S301. Weight allocation: define a weight vector and set the importance coefficient of each hardware attribute based on the degree of influence on the advertising display effect. ,in Represents the number of hardware attributes; S302, standardization processing: converting into a standardized score, for each hardware attribute , through the formula Convert it to a normalized score, where and are the mean and standard deviation of the attribute, respectively.

[0010] As a preferred technical solution of the present application, S300 specifically further includes the following steps: S303. Calculate the comprehensive index: The calculation formula for the comprehensive index is: ; S304. Threshold setting and classification: Setting comprehensive index thresholds ; User devices are divided into two categories based on the comprehensive index: High-quality equipment: comprehensive index is higher than or equal to the threshold equipment; Low-quality devices: The overall index is below the threshold equipment.

[0011] As a preferred technical solution of the present application, S500 also includes the following steps: S502. Calculation of expected traffic value: The historical price, click-through rate, exposure rate, and conversion rate of the ad space are used as the input features, and the comprehensive index of the user's device hardware data is used to calculate the expected traffic value. as output target.

[0012] As a preferred technical solution of the present application, S600 further includes the following steps: S600 specifically includes the following steps: S601, formulate decision rules: if the ratio of the expected traffic value to the target value threshold is greater than or equal to 1.2, then multiply the original traffic by the preset increase ratio to obtain the new advertising traffic; if it is less than 0.8, then multiply it by the preset decrease ratio; if it is between the two, keep the status quo; S602, decision rule formula: ,in, For new advertising traffic, is the original advertising traffic, is the preset reduction ratio, specifically 0.7, It is a preset increase ratio, specifically 1.5. If the ratio is greater than or equal to 1.2, it means that the expected traffic value is significantly higher than the target threshold, indicating that the ad slot has high potential and is worth investing more resources in. If the ratio is less than 0.8, it means that the expected traffic value is much lower than the target threshold, which may mean that the ad slot performs poorly and needs to reduce investment to avoid wasting resources. If the ratio is between 0.8 and 1.2, the expected traffic value is close to the target threshold and there is no need to make significant adjustments to the traffic for the time being. You can continue to observe for a while.

[0013] As a preferred technical solution of the present application, S700 specifically includes the following steps: S701. Dynamically adjust the advertisement content through the following model, where the model is represented as follows: , input features: comprehensive index , Current network speed , Original ad size , original quality of advertisement , Original frame rate of advertisement , output result: adjusted ad size , Adjusted ad quality , Adjusted ad frame rate , and is the model parameter.

[0014] As the preferred technical solution of this application, it includes the following modules: Data collection module: responsible for collecting historical data of advertising positions and user device hardware data; Data preprocessing module: cleans, standardizes and performs feature engineering on raw data; Traffic prediction module: predicts advertising traffic through models; Hardware index calculation module: evaluates user device performance, calculates a comprehensive index, and classifies high-quality and low-quality devices; Small traffic testing module: perform A / B testing, detect potential problems and optimize advertising content; Traffic value calculation module: calculate the expected traffic value; Decision optimization module: adjusts advertising traffic according to traffic ratio and formulates a strategy that prioritizes user experience.

[0015] Compared with the prior art, the present invention has the following beneficial effects: In the scheme of this application: 1. Comprehensively consider the historical price, click-through rate, exposure rate and conversion rate of the ad space, user device performance and other factors to build a more accurate traffic prediction model, which reduces the labor intensity and risk of manual calculations. It can not only predict traffic, but also adjust the advertising delivery strategy according to the expected value of traffic. By understanding the user's device performance information to predict the expected value of traffic, it reduces the possibility of user loss and decreased conversion rate caused by slow loading or stuck ads due to device problems, thereby optimizing advertising delivery strategies and improving user experience and conversion rates. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A schematic diagram of the advertising traffic delivery method provided for this application; Figure 2 A schematic diagram of the advertising traffic delivery device provided in this application; DETAILED DESCRIPTION In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. 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 creative work should fall within the scope of protection of the present invention.

[0017] As described in the background technology, in traditional advertising traffic delivery plans, ad slot operators manually calculate the traffic to be delivered based on the demand for advertising delivery and the estimated daily traffic of the ad slots, and formulate an advertising traffic delivery plan for the ad slots to be delivered. Usually, ad slot operators generally perform manual calculations based on their own experience or fixed formulas, which is a large workload and requires calculations every time. Not only is the workload large, but there is also a high risk of delivering advertisements with large traffic to multiple ad slots at one time. If the publicity effect is not good, the business value will be greatly lost.

[0018] In order to solve this technical problem, the present invention provides an advertisement traffic delivery method, device and storage medium, which are applied to the field of advertisement traffic delivery.

[0019] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features and technical solutions in the embodiments may be combined with each other.

[0020] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, further definition and explanation thereof is not required in subsequent drawings.

[0021] Example 1, please refer to Figure 1, an advertisement traffic delivery method, characterized in that it comprises the following steps: S100. Data collection: According to the type of advertisement to be placed (e.g., if the advertisement is a car advertisement, data of car advertisements will be collected; if the advertisement is a home appliance advertisement, data of home appliance advertisements will be collected), the historical price, click-through rate, exposure rate and conversion rate of the corresponding advertisement space are obtained from the advertising server (it should be noted that the above data are all data of the previous natural month) (it can be collected through the advertising server and third-party platforms, which is an existing public technology and will not be described in detail), and the hardware data of the target media platform users are directly collected from the application through SDK integration. The hardware data includes screen resolution, processor frequency, running memory, network speed and energy consumption data (it can be collected through the advertising server and third-party platforms, which is an existing public technology and will not be described in detail, ensuring that all data collection activities comply with privacy regulations such as GDPR, and use anonymization or pseudonymization technology to process personal information); S200, calculating the advertising traffic prediction value: using the data collected in step S100 to build an advertising traffic prediction model to predict the advertising traffic; S200 specifically includes the following steps: S201. Construct prediction function: XGBoost is used as a prediction model, and its basic form can be expressed as ,in is the predicted advertising traffic, is the input feature vector, It is the nonlinear mapping function provided by the trained XGBoost model; S202, application model: taking new ad space data and user device hardware data as input features , use the trained XGBoost model to predict the advertising traffic ; S300, calculating a comprehensive index of hardware data: calculating a comprehensive index based on the hardware data of the user's device to evaluate the display effect of the advertisement on different devices; S400, traffic test: In a real environment, a small-scale test is conducted on the selected ad slots according to the first preset traffic, and the initial test ratio is set to 5% to 10% of the total user group. An A / B test framework is created, which includes a control group and an experimental group. The experimental group applies the optimized ad delivery strategy in S600, while the control group continues to use the existing plan to obtain the test results. S500, using a traffic value model to calculate the expected traffic value: calculating the test results to obtain the expected traffic value of each advertising position; S500 specifically includes the following steps: S501, model structure: Expected value of traffic formula: ,in, is the click-through rate, is the conversion rate, It is a comprehensive index of the user's device hardware data. It's the exposure rate. is the weight coefficient ( and Specifically, it is 0.4, and the rest are 0.2, which means that the formula for expected traffic value places more emphasis on comprehensive index and conversion rate). is the random error term; S600, Traffic Ratio and Threshold Comparison and Strategy Optimization: Calculating Expected Traffic Value With target value threshold Traffic ratio (target value threshold is the historical average traffic value), and the advertising traffic is adjusted according to the traffic ratio. The calculation formula is: ; S700, Optimize advertising delivery effects: Introduce a comprehensive index. When it is detected that the user device of a certain ad slot is a low-quality device, the algorithm model is adjusted to output the optimal ad format setting; S100 specifically includes the following steps: S101. Data preprocessing: Data cleaning: remove outliers, fill in missing values, and correct erroneous records; Standardization and normalization: Standardize and normalize data of different dimensions so that they are within the same numerical range for subsequent modeling (it should be noted that this is a well-known prior art and will not be described in detail here).

[0022] Example 2 further optimizes the advertising traffic delivery method provided in Example 1. Specifically, Figure 1 As shown, in S201, the goal of XGBoost is to minimize the following loss function: ,in is the loss function of a single sample, measuring the The true labels of samples With the predicted value The difference between (such as squared loss), is a regularization term used to prevent overfitting.

[0023] Embodiment 3 further optimizes the advertising traffic delivery method provided in Embodiment 1 or 2. Specifically, Figure 1 As shown, S301, weight allocation: define a weight vector, and set the importance coefficient of each hardware attribute based on the degree of influence on the advertising display effect ,in Indicates the number of hardware attributes (screen resolution, processor frequency, running memory, network speed and energy consumption data in S100); S302, standardization processing: converting into a standardized score, for each hardware attribute , through the formula Convert it to a normalized score, where and are the mean and standard deviation of the attribute, respectively.

[0024] S300 specifically includes the following steps: S303. Calculate the comprehensive index: The calculation formula for the comprehensive index is: ; S304. Threshold setting and classification: Setting comprehensive index thresholds (Specifically 1); User devices are divided into two categories based on the comprehensive index: High-quality equipment: comprehensive index is higher than or equal to the threshold equipment; Low-quality devices: The overall index is below the threshold equipment.

[0025] Example 4, please refer to Figure 1 , S502, Calculation of expected traffic value: The historical price, click-through rate, exposure rate, and conversion rate of the advertising space are used as the comprehensive index of the user's device hardware data as input features, and the expected traffic value is calculated. as output target.

[0026] S600 further includes the following steps: S600 specifically includes the following steps: S601, formulate decision rules: if the ratio of the expected traffic value to the target value threshold is greater than or equal to 1.2, then multiply the original traffic by the preset increase ratio to obtain the new advertising traffic; if it is less than 0.8, then multiply it by the preset decrease ratio; if it is between the two, keep the status quo; S602, decision rule formula: ,in, For new advertising traffic, is the original advertising traffic, is the preset reduction ratio, specifically 0.7, It is a preset increase ratio, specifically 1.5. If the ratio is greater than or equal to 1.2, it means that the expected traffic value is significantly higher than the target threshold, indicating that the ad slot has high potential and is worth investing more resources in. If the ratio is less than 0.8, it means that the expected traffic value is much lower than the target threshold, which may mean that the ad slot performs poorly and needs to reduce investment to avoid wasting resources. If the ratio is between 0.8 and 1.2, the expected traffic value is close to the target threshold and there is no need to make significant adjustments to the traffic for the time being. You can continue to observe for a while.

[0027] Example 5, please refer to Figure 1 , S700 specifically includes the following steps: S701. Dynamically adjust the advertisement content through the following model, where the model is represented as follows: , input features: comprehensive index , Current network speed , Original ad size , original quality of advertisement , Original frame rate of advertisement , output result: adjusted ad size , Adjusted ad quality , Adjusted ad frame rate , (Indicates that the expression in S701 - )and (Indicates that the expression in S701 - ) is a model parameter. The above linear regression model can be trained using historical data to ensure that the model can accurately adjust the advertising content according to the comprehensive index and other conditions. If the user's comprehensive index is lower than the threshold (i.e., poor device performance), reduce the resolution of the ad, lower the frame rate, and appropriately compress the file size to reduce the burden on the device. User experience becomes one of the primary considerations. For low-quality devices, even if the traffic value is high, priority should be given to ensuring that the ad can be played smoothly without affecting the user experience.

[0028] S702, advertisement content adjustment algorithm; 7.2.1. Introducing Deep Reinforcement Learning (DRL); Define the state space and action space: State space: includes comprehensive index of user device, current network speed, original ad size, image quality, frame rate, etc.

[0029] Action space: A collection of actions for adjusting ad content, such as resizing, reducing image quality, or reducing the number of frames.

[0030] 7.2.2. Train the DRL model; Select model architecture: You can choose common algorithms such as Deep Q Network (DQN) and Proximal Policy Optimization (PPO).

[0031] Simulation environment: Create a simulation environment for training the DRL model, in which ad playback situations under different user device conditions are simulated.

[0032] Optimization strategy: By interacting with the simulated environment, the model gradually learns the best ad content adjustment strategy to maximize the cumulative reward.

[0033] 7.2.3. Apply the DRL model; Deployment model: Apply the trained DRL model in the actual advertising process and dynamically adjust the advertising content based on the status information collected in real time.

[0034] Continuous improvement: As more real data accumulates, the DRL model is continuously optimized to ensure that it is always in the best performance state.

[0035] S703, dynamic adaptive adjustment; Deploy a real-time monitoring system: continuously track the performance of ads on different devices and collect user feedback and interaction data.

[0036] Adaptive adjustment: Automatically adjust ad content based on real-time data feedback to ensure a good user experience even when network conditions change or user device performance fluctuates.

[0037] S704, user experience priority sorting; Make rules: Make user experience one of the top considerations. For example, for low-quality devices, even if the traffic value is higher, it should be prioritized to ensure that ads can be played smoothly without affecting the user experience.

[0038] Example 6, please refer to Figure 2 , an advertising traffic delivery device, comprising the following modules: Data collection module: responsible for collecting historical data of advertising positions and user device hardware data; Data preprocessing module: cleans, standardizes and performs feature engineering on raw data; Traffic prediction module: predicts advertising traffic through models; Hardware index calculation module: evaluates user device performance, calculates a comprehensive index, and classifies high-quality and low-quality devices; Small traffic testing module: perform A / B testing, detect potential problems and optimize advertising content; Traffic value calculation module: calculate the expected traffic value; Decision optimization module: adjusts advertising traffic according to traffic ratio and formulates a strategy that prioritizes user experience.

[0039] In the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", "connected", "fixed" and the like should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral one; it can be a mechanical connection, an electrical connection, or communication with each other; it can be a direct connection, or an indirect connection through an intermediate medium, it can be the internal connection of two elements or the interaction relationship between two elements, unless otherwise clearly defined. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0040] Obviously, the embodiments described above are only some embodiments of the present invention, rather than all embodiments. The preferred embodiments of the present invention are given in the accompanying drawings, but they do not limit the patent scope of the present invention. The present invention can be implemented in many different forms. On the contrary, the purpose of providing these embodiments is to make the understanding of the disclosure of the present invention more thorough and comprehensive. Although the present invention has been described in detail with reference to the aforementioned embodiments, for those skilled in the art, it is still possible to modify the technical solutions recorded in the aforementioned specific embodiments, or to perform equivalent replacements for some of the technical features therein. Any equivalent structure made using the contents of the specification and drawings of the present invention, directly or indirectly used in other related technical fields, is also within the scope of patent protection of the present invention.

Claims

1. A method for delivering advertising traffic, characterized in that: The following steps are involved: S100, data collection: according to the type of advertisements to be placed, the historical price, click-through rate, exposure rate and conversion rate of the corresponding advertisement slots are obtained from the advertisement server. The hardware data of the target media platform users are directly collected from the application through SDK integration. The hardware data includes screen resolution, processor frequency, running memory, network speed and energy consumption data. S200, calculating the advertising traffic prediction value: using the data collected in step S100 to build an advertising traffic prediction model to predict the advertising traffic; S200 specifically includes the following steps: S201. Construct prediction function: XGBoost is used as a prediction model, and its basic form can be expressed as ,in is the predicted advertising traffic, is the input feature vector, It is the nonlinear mapping function provided by the trained XGBoost model; S202, application model: taking new ad space data and user device hardware data as input features , predict the advertising traffic through the trained XGBoost model ; S300, calculating a comprehensive index of hardware data: calculating a comprehensive index based on the hardware data of the user's device to evaluate the display effect of the advertisement on different devices; S400, traffic test: conduct a small-scale test on the selected ad slot in a real environment according to the first preset traffic, set the initial test ratio to 5% to 10% of the total user group, create an A / B test framework, and respectively include a control group and an experimental group, and obtain the test results; S500, using a traffic value model to calculate the expected traffic value: calculating the test results to obtain the expected traffic value of each advertising position; S500 specifically includes the following steps: S501, model structure: Expected value of traffic formula: ,in, is the click-through rate, is the conversion rate, It is a comprehensive index of the user's device hardware data. It's the exposure rate. is the weight coefficient, is the random error term; S600, Traffic Ratio and Threshold Comparison and Strategy Optimization: Calculating Expected Traffic Value With target value threshold The traffic ratio is calculated by adjusting the advertising traffic according to the traffic ratio. The calculation formula is: ; S700. Optimize advertising delivery effects: Introduce a comprehensive index. When it is detected that the user device of a certain ad slot is a low-quality device, the algorithm model is adjusted to output the optimal ad format settings.

2. The method for delivering advertisement traffic according to claim 1, characterized in that: S100 specifically includes the following steps: S101. Data preprocessing: Data cleaning: remove outliers, fill in missing values, and correct erroneous records; Standardization and normalization: Standardize and normalize data of different dimensions so that they are in the same numerical range for subsequent modeling.

3. The method for delivering advertisement traffic according to claim 2, characterized in that: In S201, the goal of XGBoost is to minimize the following loss function: ,in is the loss function of a single sample, measuring the The true labels of samples With the predicted value The difference between is a regularization term used to prevent overfitting.

4. The method for delivering advertisement traffic according to claim 3, characterized in that: S300 specifically includes the following steps: S301. Weight allocation: define a weight vector and set the importance coefficient of each hardware attribute based on the degree of influence on the advertising display effect. ,in Represents the number of hardware attributes; S302, standardization processing: converting into a standardized score, for each hardware attribute , through the formula Convert it to a normalized score, where and are the mean and standard deviation of the attribute, respectively.

5. The method for delivering advertisement traffic according to claim 4, characterized in that: S300 specifically includes the following steps: S303. Calculate the comprehensive index: The calculation formula for the comprehensive index is: ; S304. Threshold setting and classification: Setting comprehensive index thresholds ; User devices are divided into two categories based on the comprehensive index: High-quality equipment: comprehensive index is higher than or equal to the threshold equipment; Low-quality devices: The overall index is below the threshold equipment.

6. The method for delivering advertisement traffic according to claim 5, characterized in that: S500 also includes the following steps: S502. Calculation of expected traffic value: The historical price, click-through rate, exposure rate, and conversion rate of the ad space are used as the input features, and the comprehensive index of the user's device hardware data is used to calculate the expected traffic value. as output target.

7. The method for delivering advertisement traffic according to claim 6, characterized in that: S600 further includes the following steps: S600 specifically includes the following steps: S601, formulate decision rules: if the ratio of the expected traffic value to the target value threshold is greater than or equal to 1.2, then multiply the original traffic by the preset increase ratio to obtain the new advertising traffic; if it is less than 0.8, then multiply it by the preset decrease ratio; if it is between the two, keep the status quo; S602, decision rule formula: ,in, For new advertising traffic, is the original advertising traffic, is the preset reduction ratio, specifically 0.7, It is the preset increase ratio, which is 1.

5.

8. The method for delivering advertisement traffic according to claim 1, characterized in that: S700 specifically includes the following steps: S701. Dynamically adjust the advertisement content through the following model, where the model is represented as follows: , input features: comprehensive index , Current network speed , Original ad size , original quality of advertisement , Original frame rate of advertisement , output result: adjusted ad size , Adjusted ad quality , Adjusted ad frame rate , and is the model parameter.

9. An advertisement traffic delivery device, using the advertisement traffic delivery method as claimed in claim 1, characterized in that: Includes the following modules: Data collection module: responsible for collecting historical data of advertising positions and user device hardware data; Data preprocessing module: cleans, standardizes and performs feature engineering on raw data; Traffic prediction module: predicts advertising traffic through models; Hardware index calculation module: evaluates user device performance, calculates a comprehensive index, and classifies high-quality and low-quality devices; Small traffic testing module: perform A / B testing, detect potential problems and optimize advertising content; Traffic value calculation module: calculate the expected traffic value; Decision optimization module: adjusts advertising traffic according to traffic ratio and formulates a strategy that prioritizes user experience.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps of the method according to any one of claims 1 to 8 are implemented.

Images