Personalized advertisement pushing system based on big data and artificial intelligence

Abstract

The invention belongs to the technical field of advertisement pushing, and particularly relates to a personalized advertisement pushing system based on big data and artificial intelligence, and the system comprises a data collection module which is used for collecting user data from various channels, and the user data comprises basic information, browsing records, purchase history, search behaviors and social interaction of users; and the data storage and preprocessing module is used for storing the collected data in a distributed database, carrying out cleaning, duplicate removal and standardized preprocessing operations so as to ensure the quality and availability of the data, and classifying and marking the data at the same time. According to the method, the multi-dimensional data of the user is collected through big data, and the user portrait model and the advertisement effect prediction model are constructed by applying a machine learning algorithm, so that the matching degree of the advertisement and the interest of the user can be greatly improved, the click rate of the accurately pushed advertisement is improved to 5% or above from about 0.1% in the prior art, and the problem of low advertisement accuracy is effectively solved.

Description

Technical Field

[0001] This invention relates to the field of advertising push technology, specifically a personalized advertising push system based on big data and artificial intelligence. Background Technology

[0002] In today's digital age, with the widespread adoption of the internet and mobile devices, the advertising industry is undergoing profound changes. Traditional advertising methods, such as designing and distributing ad content based on fixed rules or experience, are no longer sufficient to attract consumer attention or meet the diverse needs of modern consumers. According to relevant data, consumers are exposed to an average of over 5,000 advertising messages every day, but the vast majority of these are ignored or directly blocked. This not only leads to a waste of advertising resources but also makes it difficult for advertisers to achieve their desired advertising results.

[0003] Traditional advertising methods have the following problems:

[0004] 1. Low Ad Targeting: Traditional advertising models struggle to accurately target the needs and preferences of different users. For example, on an e-commerce platform, user A frequently browses and purchases sports equipment, while user B is more inclined to buy beauty products. However, traditional advertising methods might push beauty ads to user A and sports equipment ads to user B, resulting in a mismatch between ads and user interests and poor advertising effectiveness. Statistics show that the average click-through rate (CTR) of traditional advertising is only around 0.1%, while the CTR of precisely targeted ads can reach over 5%.

[0005] 2. Poor User Experience: When ad content doesn't align with user interests, it easily provokes user resentment and degrades the user experience. For example, when browsing news apps, users frequently see ads unrelated to their interests, which interferes with normal app usage and may even lead to uninstallation. A survey of app users showed that over 60% of users reported uninstalling apps due to poor ad experiences.

[0006] 3. Difficulty in Measuring Advertising Effectiveness: The lack of effective data analysis and evaluation methods makes it difficult to accurately assess the effectiveness of advertising campaigns. Advertisers often cannot clearly understand how many potential customers were attracted or how many ultimately made a purchase. This makes it difficult for advertisers to adjust and optimize their advertising strategies based on results. For example, after an advertising campaign, a company may not be able to accurately determine the specific contribution of the advertising to brand awareness, product sales growth, etc.

[0007] 4. Significant resource waste: Due to low ad targeting accuracy, a large amount of advertising resources are wasted on users who are not interested in the ads. For example, a brand spends a lot of money on advertising on social media platforms, but because the ad targeting is inaccurate, only a small portion of the target customers see the ads, and most ad impressions do not generate any real value.

[0008] Based on the above, a personalized advertising push system based on big data and artificial intelligence is invented. Summary of the Invention

[0009] To address the aforementioned technical problems, according to one aspect of the present invention, the present invention provides the following technical solution:

[0010] Personalized advertising push systems based on big data and artificial intelligence include:

[0011] The data collection module is used to collect user data from various channels, including basic user information, browsing history, purchase history, search behavior, and social interactions.

[0012] The data storage and preprocessing module is used to store the collected data in a distributed database and perform cleaning, deduplication, and standardization preprocessing operations to ensure data quality and availability. At the same time, it classifies and labels the data to prepare for subsequent data analysis and model training.

[0013] The data analysis and modeling module is used to analyze the preprocessed data and build corresponding models through machine learning algorithms. The models include user profile models and advertising effect prediction models.

[0014] The intelligent ad creative generation module is used to automatically generate diverse ad creatives based on user profiles, ad placement goals, and current market trend information using artificial intelligence technology, for the ad push decision module to select and use.

[0015] The ad push decision module is used to formulate ad push strategies based on the results of the data analysis and modeling module and the intelligent ad creative generation module, combined with the advertiser's placement needs and budget.

[0016] The traffic control module is used to first obtain real-time information on the traffic situation of each advertising channel, the distribution of user access time periods, and the speed at which advertisers' budgets are consumed. Then, based on preset rules and algorithms, it adjusts the advertising strategy formulated by the advertising push decision-making layer at the traffic level, and then passes the optimized instructions to the advertising execution module.

[0017] The ad delivery execution module is used to push ads to target users, supports multiple ad formats, and monitors the ad delivery effect in real time, collects user feedback data, and feeds the data back to the data analysis and modeling module to optimize the model and strategy.

[0018] As a preferred embodiment of the personalized advertising push system based on big data and artificial intelligence described in this invention, the data analysis and modeling module includes:

[0019] The data import and integration module is used to obtain cleaned, deduplicated, and standardized user data from the data storage and preprocessing module, and to integrate data from different sources to form a complete dataset, preparing for subsequent analysis and modeling.

[0020] The feature engineering module is used to extract features from the integrated data. By analyzing the characteristics of the data and business requirements, it extracts variables that can effectively describe user behavior and characteristics.

[0021] The model selection and training module is used to select the corresponding machine learning algorithms and models based on the data analysis objectives and data characteristics. The machine learning algorithms include classification algorithms, clustering algorithms, and recommendation algorithms. At the same time, it can select the corresponding model for training for each specific business problem. During the training process, the dataset is divided into training set and test set, and the training set is used to train the model.

[0022] As a preferred embodiment of the personalized advertising push system based on big data and artificial intelligence described in this invention, the data analysis and modeling module further includes:

[0023] The model evaluation and optimization module is used to evaluate the trained model using a test set, and to determine whether the model's performance meets business requirements based on the evaluation results.

[0024] The model deployment and update module is used to deploy the optimized model to the production environment and integrate it with the ad push decision module to realize personalized ad push. At the same time, during actual operation, new data is continuously collected, and the model is regularly updated and optimized according to business needs and data changes to ensure the accuracy and timeliness of the model.

[0025] As a preferred embodiment of the personalized advertising push system based on big data and artificial intelligence described in this invention, the intelligent advertising creative generation module includes:

[0026] The data acquisition and integration module is used to acquire user data from the data collection module, and at the same time receive advertising materials, campaign objectives and budget information uploaded by advertisers, and integrate the acquired data to form the basic dataset for creative generation, providing comprehensive data support for subsequent analysis;

[0027] The user characteristics and needs analysis module is used to conduct in-depth analysis of the integrated data using natural language processing, image recognition and artificial intelligence technologies. At the same time, it combines the advertiser's campaign goals to accurately locate the characteristics and needs of the target user group and provide direction for creative generation.

[0028] The creative element generation module is used to automatically generate advertising creative elements based on the results of the user characteristics and needs analysis module, using generative adversarial networks and variational autoencoder deep learning models.

[0029] As a preferred embodiment of the personalized advertising push system based on big data and artificial intelligence described in this invention, the intelligent advertising creative generation module further includes:

[0030] The Creative Combination and Optimization module is used to combine various generated creative elements to form multiple advertising creative schemes with different styles and versions. At the same time, the algorithm is used to evaluate the schemes, scoring them based on the appeal, relevance, and uniqueness of the creatives, and selecting the high-scoring schemes. In addition, the creative schemes are optimized and adjusted based on user feedback data and historical data on advertising performance.

[0031] The output and push module is used to output the optimized advertising creative plan to the advertising push decision module, so that it can select the corresponding creative for advertising based on user profile and advertising strategy factors.

[0032] As a preferred embodiment of the personalized advertising push system based on big data and artificial intelligence described in this invention, the advertising push decision module includes:

[0033] The receiving data and information integration module is used to first obtain relevant data from the data analysis and modeling module and the intelligent advertising creative generation module, and at the same time receive the placement requirements and advertising material information set by the advertiser. Then, the data is integrated to build a basic information database for advertising push decisions, providing a comprehensive basis for subsequent decisions.

[0034] The initial strategy formulation module is used to formulate an initial advertising push strategy based on the integrated information and using preset decision-making algorithms and rules.

[0035] The strategy evaluation and optimization module is used to evaluate the initially formulated advertising push strategy from multiple dimensions. Through simulated deployment, it analyzes the possible effects of the strategy in different scenarios, and optimizes and adjusts the strategy when the evaluation results do not meet expectations.

[0036] As a preferred embodiment of the personalized advertising push system based on big data and artificial intelligence described in this invention, the advertising push decision module further includes:

[0037] The strategy review and finalization module is used to submit the optimized advertising push strategy to the reviewers or review system for review, check whether the strategy complies with laws and regulations, platform rules and advertiser requirements. After the review is approved, the final advertising push strategy is determined and passed to the advertising execution module so that the advertising can be carried out. If the review fails, the module returns to the previous step to continue to modify and optimize the strategy.

[0038] The strategy dynamic monitoring and adjustment module is used to monitor the advertising performance data in real time during the advertising campaign and compare it with the expected results. Once a deviation between the actual results and the expectations is found, or if the market environment or user behavior changes, the advertising push strategy can be dynamically adjusted in a timely manner to ensure that the advertising campaign always moves towards the best results.

[0039] As a preferred embodiment of the personalized advertising push system based on big data and artificial intelligence described in this invention, the traffic control module includes:

[0040] The real-time traffic monitoring module is used to collect traffic data from various advertising channels in real time, and at the same time, obtain the advertiser's real-time budget consumption data and the advertising performance data fed back by the advertising execution module.

[0041] The traffic data analysis module is used to perform in-depth analysis of the collected traffic and advertising data, and to use statistical methods and data mining techniques to find the relationship between traffic change patterns, user behavior patterns and advertising effectiveness.

[0042] The traffic control strategy formulation module is used to formulate traffic control strategies based on data analysis results, combined with the advertiser's campaign goals and remaining budget.

[0043] As a preferred embodiment of the personalized advertising push system based on big data and artificial intelligence described in this invention, the traffic control module further includes:

[0044] The control instruction issuance module is used to convert the formulated traffic control strategy into specific instructions and send them to the advertising execution module.

[0045] The performance feedback and strategy optimization module is used to continuously track the advertising performance after traffic control, collect new traffic and campaign data, compare key indicators before and after control, and evaluate the effectiveness of the control strategy.

[0046] Compared with existing technologies:

[0047] 1. Addressing the issue of low advertising accuracy: This invention collects multi-dimensional user data through big data and uses machine learning algorithms to build user profile models and advertising effect prediction models. Compared with traditional advertising methods, it can significantly improve the matching degree between advertisements and user interests, increasing the click-through rate of targeted advertisements from about 0.1% to more than 5%, effectively solving the problem of low advertising accuracy.

[0048] 2. Addressing poor user experience: This invention comprehensively considers factors such as user interests, historical behavior, and current scenario, as well as indicators like ad relevance, quality, and delivery effectiveness for ad recommendations. When users browse news and information apps, the system can accurately push ads that match their interests, reducing irrelevant ad interference and preventing users from uninstalling apps due to poor ad experiences, thus improving the user experience when using various applications.

[0049] 3. Addressing the problem of difficulty in measuring advertising effectiveness: This invention monitors and analyzes metrics such as ad exposure, click-through rate, conversion rate, and purchase volume in real time. This allows advertisers to clearly understand the number of potential customers attracted and actual purchasing behavior after ad campaigns, thereby adjusting and optimizing their ad campaign strategies based on the data, thus solving the problem of difficulty in measuring advertising effectiveness.

[0050] 4. Addressing resource waste: This invention utilizes big data analytics and artificial intelligence to accurately target customers. When advertisers place ads on social media platforms, the system can use user profiles and predictive models to push ads to genuinely interested user groups, avoiding the waste of significant advertising resources on non-target customers, improving advertising resource utilization efficiency, and reducing advertising costs. Attached Figure Description

[0051] Figure 1 This is a schematic diagram of the overall framework of the present invention. Detailed Implementation

[0052] To make the objectives, technical solutions, and advantages of the present invention clearer, the embodiments of the present invention will be described in further detail below with reference to the accompanying drawings.

[0053] This invention provides a personalized advertising push system based on big data and artificial intelligence. Please refer to [link / reference]. Figure 1 ;

[0054] The system includes: a data collection module for collecting user data from various channels, including basic user information, browsing history, purchase history, search behavior, and social interactions; a data storage and preprocessing module for storing the collected data in a distributed database and performing preprocessing operations such as cleaning, deduplication, and standardization to ensure data quality and usability, while also classifying and labeling the data to prepare for subsequent data analysis and model training; a data analysis and modeling module for analyzing the preprocessed data and building corresponding models using machine learning algorithms, including user profile models and advertising effectiveness prediction models; and an intelligent advertising creative generation module for automatically generating diverse creative ideas based on user profiles, advertising objectives, and current market trends using artificial intelligence technology. The system generates ad creatives for the ad push decision module to select and use. The ad push decision module, based on the results from the data analysis and modeling module and the intelligent ad creative generation module, and considering the advertiser's needs and budget, formulates ad push strategies. The traffic control module first obtains real-time traffic data from various channels, user access time distribution, and advertiser budget consumption speed. Then, based on preset rules and algorithms, it adjusts the traffic-level strategy formulated by the ad push decision layer and transmits the optimized instructions to the ad execution module. The ad execution module pushes ads to target users, supports multiple ad formats, monitors ad performance in real-time, collects user feedback data, and feeds this data back to the data analysis and modeling module for model and strategy optimization.

[0055] The data analysis and modeling module includes: a data import and integration module, used to acquire cleaned, deduplicated, and standardized user data from the data storage and preprocessing module, and integrate data from different sources to form a complete dataset, preparing for subsequent analysis and modeling; a feature engineering module, used to extract features from the integrated data, extracting variables that can effectively describe user behavior and characteristics by analyzing the characteristics of the data and business needs; and a model selection and training module, used to select corresponding machine learning algorithms and models according to the data analysis objectives and data characteristics. The machine learning algorithms include classification algorithms, clustering algorithms, and recommendation algorithms. It can select the appropriate model for each specific business problem and train it accordingly. During training, the dataset is divided into training and test sets, and the training set is used to train the model. The model evaluation and optimization module is used to evaluate the trained model using the test set and determine whether the model's performance meets business requirements based on the evaluation results. The model deployment and update module is used to deploy the optimized model to the production environment and integrate it with the ad push decision module to achieve personalized ad push. At the same time, during actual operation, new data is continuously collected, and the model is regularly updated and optimized according to business needs and data changes to ensure the accuracy and timeliness of the model.

[0056] The intelligent advertising creative generation module includes: a data acquisition and integration module, used to acquire user data from the data collection module, and simultaneously receive advertising materials, targeting objectives, and budget information uploaded by advertisers, and integrate the acquired data to form a basic dataset for creative generation, providing comprehensive data support for subsequent analysis; a user characteristic and demand analysis module, used to perform in-depth analysis of the integrated data using natural language processing, image recognition, and artificial intelligence technologies, and combined with the advertiser's targeting objectives to accurately locate the characteristics and needs of the target user group, providing direction for creative generation; and a creative element generation module, used to generate creative elements based on the results of the user characteristic and demand analysis module. The system automatically generates ad creative elements using generative adversarial networks and variational autoencoder deep learning models. The creative combination and optimization module combines these elements to create multiple ad creative schemes with different styles and versions. Algorithms evaluate these schemes, scoring them based on attractiveness, relevance, and uniqueness, selecting the highest-scoring schemes. Furthermore, the system optimizes and adjusts the creative schemes based on user feedback data and historical ad performance data. The output and push module delivers the optimized ad creative schemes to the ad push decision module, which then combines user profiles and ad delivery strategy factors to select the appropriate creative for ad delivery.

[0057] By setting up an intelligent ad creative generation module, it can solve the problem of monotonous ad creatives that are difficult to attract users. For example, it can intelligently generate trendy and personalized ad copy and visual designs for young and fashionable users; for family users, it can generate warm and practical ad content. By providing rich and personalized ad creatives, it can improve the attractiveness and competitiveness of ads, thereby increasing click-through rates and exposure, while reducing the burden on advertisers to create ad creatives and improving ad delivery efficiency.

[0058] The ad push decision-making module includes: a data receiving and information integration module, used to first obtain relevant data from the data analysis and modeling module and the intelligent ad creative generation module, and simultaneously receive the advertiser's placement requirements and ad creative information, then integrate the data to build a basic information database for ad push decisions, providing a comprehensive basis for subsequent decisions; a preliminary strategy formulation module, used to formulate a preliminary ad push strategy based on the integrated information and using preset decision algorithms and rules; a strategy evaluation and optimization module, used to evaluate the preliminary ad push strategy from multiple dimensions, analyze the possible effects of the strategy in different scenarios through simulated placement, and optimize and adjust the strategy when the evaluation results do not meet expectations; and a strategy review and... The finalization module submits the optimized ad push strategy to reviewers or the review system for approval. It checks whether the strategy complies with laws, regulations, platform rules, and advertiser requirements. If approved, the final ad push strategy is determined and passed to the ad execution module for ad campaign implementation. If the review fails, the module returns to the previous step to continue modifying and optimizing the strategy. The dynamic monitoring and adjustment module monitors ad performance data in real time during ad execution and compares it with expected results. If deviations are found between actual and expected results, or if market conditions or user behavior change, the module dynamically adjusts the ad push strategy to ensure the ad campaign consistently achieves optimal results.

[0059] The traffic control module includes: a real-time traffic monitoring module, used to collect traffic data from various advertising channels in real time, and simultaneously obtain real-time budget consumption data of advertisers, as well as performance data fed back by the advertising execution module; a traffic data analysis module, used to perform in-depth analysis of the collected traffic and performance data, using statistical methods and data mining techniques to identify the relationship between traffic change patterns, user behavior patterns, and advertising performance; a control strategy formulation module, used to formulate traffic control strategies based on the data analysis results, combined with the advertiser's performance goals and remaining budget; a control instruction issuance module, used to convert the formulated traffic control strategies into specific instructions and send them to the advertising execution module; and an effect feedback and strategy optimization module, used to continuously track the advertising performance after traffic control, collect new traffic and performance data, and compare key indicators before and after control to evaluate the effectiveness of the control strategy.

[0060] By setting up a traffic control module, the problem of uneven ad delivery can be effectively avoided. For example, during peak periods, if a large number of users flood a particular ad channel, the traffic control module can dynamically increase the ad volume on that channel, fully utilizing peak traffic to boost ad exposure. Conversely, when a channel has insufficient traffic or a low conversion rate, the module can promptly reduce ad delivery and allocate resources to higher-quality channels, thereby improving ad resource utilization, reducing costs, and ensuring the stability of ad performance. Furthermore, by rationally controlling the ad delivery pace, it is possible to avoid over-deploying ads and disrupting user experience, thus enhancing the overall user experience.

[0061] In practical use, the specific operating steps for those skilled in the art are as follows:

[0062] S1: Collects user data from various channels through the data collection module;

[0063] S2: The data storage and preprocessing module stores the collected data in a distributed database and performs cleaning, deduplication, and standardization preprocessing operations to ensure data quality and availability. At the same time, the data is classified and labeled to prepare for subsequent data analysis and model training.

[0064] S3: The data import and integration module retrieves cleaned, deduplicated, and standardized user data from the data storage and preprocessing module. It then integrates data from different sources to form a complete dataset, preparing for subsequent analysis and modeling. After integration, the feature engineering module extracts features from the integrated data. By analyzing the data's characteristics and business needs, it extracts variables that effectively describe user behavior and characteristics. Following extraction, the model selection and training module selects appropriate machine learning algorithms and models based on the data analysis objectives and data characteristics. Furthermore, it can select corresponding models for each specific business problem. The model is trained by dividing the dataset into training and test sets. The training set is used to train the model. After training, the model evaluation and optimization module evaluates the trained model using the test set. Based on the evaluation results, it is determined whether the model's performance meets business requirements. After evaluation, the optimized model is deployed to the production environment through the model deployment and update module and integrated with the ad push decision module to achieve personalized ad push. At the same time, new data is continuously collected during actual operation. The model is updated and optimized regularly according to business needs and data changes to ensure the accuracy and timeliness of the model.

[0065] S4: The data acquisition and integration module obtains user data from the data collection module, and simultaneously receives advertising materials, campaign objectives, and budget information uploaded by advertisers. It integrates the acquired data to form the foundational dataset for creative generation, providing comprehensive data support for subsequent analysis. After integration, the user characteristic and demand analysis module utilizes natural language processing, image recognition, and artificial intelligence technologies to conduct in-depth analysis of the integrated data. Combined with the advertiser's campaign objectives, it accurately identifies the characteristics and needs of the target user group, providing direction for creative generation. Following the analysis, the creative element generation module, based on the results from the user characteristic and demand analysis module, utilizes generative adversarial networks... The system uses a deep learning model with a network and variational autoencoder to automatically generate advertising creative elements. After generation, the creative combination and optimization module combines the various creative elements to form multiple advertising creative schemes with different styles and versions. At the same time, the algorithm evaluates the schemes, scoring them based on their attractiveness, relevance, and uniqueness, and selecting the highest-scoring schemes. In addition, the creative schemes are optimized and adjusted based on user feedback data and historical data on advertising performance. After optimization, the optimized advertising creative schemes are output to the advertising push decision module through the output and push module, allowing it to combine user profiles and advertising strategy factors to select the corresponding creatives for advertising.

[0066] S5: The data and information integration module first obtains relevant data from the data analysis and modeling module and the intelligent advertising creative generation module. It also receives the advertiser's placement requirements and advertising material information. The data is then integrated to build a foundational information database for advertising push decisions, providing a comprehensive basis for subsequent decisions. After construction, the preliminary strategy formulation module uses the integrated information and preset decision-making algorithms and rules to formulate a preliminary advertising push strategy. Following this, the strategy evaluation and optimization module conducts a multi-dimensional evaluation of the preliminary advertising push strategy. Through simulated placement, it analyzes the potential effects of the strategy in different scenarios and optimizes the strategy if the evaluation results do not meet expectations. Finally, the strategy review... The finalization and confirmation module submits the optimized ad push strategy to reviewers or the review system for review. It checks whether the strategy complies with laws and regulations, platform rules, and advertiser requirements. After the review is approved, the final ad push strategy is determined and passed to the ad execution module to carry out ad delivery. If the review fails, the module returns to the previous step to continue modifying and optimizing the strategy. Once confirmed, the strategy dynamic monitoring and adjustment module can monitor the ad delivery performance data in real time during the ad delivery process and compare and analyze it with the expected results. Once a deviation between the actual results and expectations is found, or if the market environment or user behavior changes, the ad push strategy can be dynamically adjusted in a timely manner to ensure that the ad delivery always moves towards the best results.

[0067] S6: The real-time traffic monitoring module collects traffic data from various advertising channels in real time. Simultaneously, it acquires advertisers' real-time budget consumption data and advertising performance data from the advertising execution module. After acquisition, the traffic data analysis module performs in-depth analysis of the collected traffic and performance data, using statistical methods and data mining techniques to identify patterns in traffic changes, user behavior, and advertising performance. Following analysis, the control strategy formulation module, based on the data analysis results and the advertiser's goals and remaining budget, formulates a traffic control strategy. Once formulated, the control instruction issuance module translates the strategy into specific instructions and sends them to the advertising execution module. After sending, the performance feedback and strategy optimization module continuously tracks the advertising performance after traffic control, collects new traffic and performance data, compares key indicators before and after control, and evaluates the effectiveness of the control strategy.

[0068] S7: The ad delivery execution module pushes ads to target users, supports multiple ad formats, and monitors the ad delivery effect in real time, collects user feedback data, and feeds the data back to the data analysis and modeling module to optimize the model and strategy.

[0069] Although the present invention has been described above with reference to embodiments, various modifications can be made and components can be replaced with equivalents without departing from the scope of the invention. In particular, as long as there is no structural conflict, the features in the disclosed embodiments can be combined with each other in any manner. The lack of an exhaustive description of these combinations in this specification is merely for the sake of brevity and resource conservation. Therefore, the present invention is not limited to the specific embodiments disclosed herein, but includes all technical solutions falling within the scope of the claims.

Claims

1. A personalized advertising push system based on big data and artificial intelligence, characterized in that: include: The data collection module is used to collect user data from various channels, including basic user information, browsing history, purchase history, search behavior, and social interactions. The data storage and preprocessing module is used to store the collected data in a distributed database and perform cleaning, deduplication, and standardization preprocessing operations to ensure data quality and availability. At the same time, it classifies and labels the data to prepare for subsequent data analysis and model training. The data analysis and modeling module is used to analyze the preprocessed data and build corresponding models through machine learning algorithms. The models include user profile models and advertising effect prediction models. The intelligent ad creative generation module is used to automatically generate diverse ad creatives based on user profiles, ad placement goals, and current market trend information using artificial intelligence technology, for the ad push decision module to select and use. The ad push decision module is used to formulate ad push strategies based on the results of the data analysis and modeling module and the intelligent ad creative generation module, combined with the advertiser's placement needs and budget. The traffic control module is used to first obtain real-time information on the traffic situation of each advertising channel, the distribution of user access time periods, and the speed at which advertisers' budgets are consumed. Then, based on preset rules and algorithms, it adjusts the advertising strategy formulated by the advertising push decision-making layer at the traffic level, and then passes the optimized instructions to the advertising execution module. The ad delivery execution module is used to push ads to target users, supports multiple ad formats, and monitors the ad delivery effect in real time, collects user feedback data, and feeds the data back to the data analysis and modeling module to optimize the model and strategy.

2. The personalized advertising push system based on big data and artificial intelligence according to claim 1, characterized in that, The data analysis and modeling module includes: The data import and integration module is used to obtain cleaned, deduplicated, and standardized user data from the data storage and preprocessing module, and to integrate data from different sources to form a complete dataset, preparing for subsequent analysis and modeling. The feature engineering module is used to extract features from the integrated data. By analyzing the characteristics of the data and business requirements, it extracts variables that can effectively describe user behavior and characteristics. The model selection and training module is used to select the corresponding machine learning algorithms and models based on the data analysis objectives and data characteristics. The machine learning algorithms include classification algorithms, clustering algorithms, and recommendation algorithms. At the same time, it can select the corresponding model for training for each specific business problem. During the training process, the dataset is divided into training set and test set, and the training set is used to train the model.

3. The personalized advertising push system based on big data and artificial intelligence according to claim 2, characterized in that, The data analysis and modeling module also includes: The model evaluation and optimization module is used to evaluate the trained model using a test set, and to determine whether the model's performance meets business requirements based on the evaluation results. The model deployment and update module is used to deploy the optimized model to the production environment and integrate it with the ad push decision module to realize personalized ad push. At the same time, during actual operation, new data is continuously collected, and the model is regularly updated and optimized according to business needs and data changes to ensure the accuracy and timeliness of the model.

4. The personalized advertising push system based on big data and artificial intelligence according to claim 1, characterized in that, The intelligent advertising creative generation module includes: The data acquisition and integration module is used to acquire user data from the data collection module, and at the same time receive advertising materials, campaign objectives and budget information uploaded by advertisers, and integrate the acquired data to form the basic dataset for creative generation, providing comprehensive data support for subsequent analysis; The user characteristics and needs analysis module is used to conduct in-depth analysis of the integrated data using natural language processing, image recognition and artificial intelligence technologies. At the same time, it combines the advertiser's campaign goals to accurately locate the characteristics and needs of the target user group and provide direction for creative generation. The creative element generation module is used to automatically generate advertising creative elements based on the results of the user characteristics and needs analysis module, using generative adversarial networks and variational autoencoder deep learning models.

5. The personalized advertising push system based on big data and artificial intelligence according to claim 4, characterized in that, The intelligent advertising creative generation module also includes: The Creative Combination and Optimization module is used to combine various generated creative elements to form multiple advertising creative schemes with different styles and versions. At the same time, the algorithm is used to evaluate the schemes, scoring them based on the appeal, relevance, and uniqueness of the creatives, and selecting the high-scoring schemes. In addition, the creative schemes are optimized and adjusted based on user feedback data and historical data on advertising performance. The output and push module is used to output the optimized advertising creative plan to the advertising push decision module, so that it can select the corresponding creative for advertising based on user profile and advertising strategy factors.

6. The personalized advertising push system based on big data and artificial intelligence according to claim 1, characterized in that, The ad push decision module includes: The receiving data and information integration module is used to first obtain relevant data from the data analysis and modeling module and the intelligent advertising creative generation module, and at the same time receive the placement requirements and advertising material information set by the advertiser. Then, the data is integrated to build a basic information database for advertising push decisions, providing a comprehensive basis for subsequent decisions. The initial strategy formulation module is used to formulate an initial advertising push strategy based on the integrated information and using preset decision-making algorithms and rules. The strategy evaluation and optimization module is used to evaluate the initially formulated advertising push strategy from multiple dimensions. Through simulated deployment, it analyzes the possible effects of the strategy in different scenarios, and optimizes and adjusts the strategy when the evaluation results do not meet expectations.

7. The personalized advertising push system based on big data and artificial intelligence according to claim 6, characterized in that, The ad push decision module also includes: The strategy review and finalization module is used to submit the optimized advertising push strategy to the reviewers or review system for review, check whether the strategy complies with laws and regulations, platform rules and advertiser requirements. After the review is approved, the final advertising push strategy is determined and passed to the advertising execution module so that the advertising can be carried out. If the review fails, the module returns to the previous step to continue to modify and optimize the strategy. The strategy dynamic monitoring and adjustment module is used to monitor the advertising performance data in real time during the advertising campaign and compare it with the expected results. Once a deviation between the actual results and the expectations is found, or if the market environment or user behavior changes, the advertising push strategy can be dynamically adjusted in a timely manner to ensure that the advertising campaign always moves towards the best results.

8. The personalized advertising push system based on big data and artificial intelligence according to claim 1, characterized in that, The flow control module includes: The real-time traffic monitoring module is used to collect traffic data from various advertising channels in real time, and at the same time, obtain the advertiser's real-time budget consumption data and the advertising performance data fed back by the advertising execution module. The traffic data analysis module is used to perform in-depth analysis of the collected traffic and advertising data, and to use statistical methods and data mining techniques to find the relationship between traffic change patterns, user behavior patterns and advertising effectiveness. The traffic control strategy formulation module is used to formulate traffic control strategies based on data analysis results, combined with the advertiser's campaign goals and remaining budget.

9. The personalized advertising push system based on big data and artificial intelligence according to claim 8, characterized in that, The flow control module also includes: The control instruction issuance module is used to convert the formulated traffic control strategy into specific instructions and send them to the advertising execution module. The performance feedback and strategy optimization module is used to continuously track the advertising performance after traffic control, collect new traffic and campaign data, compare key indicators before and after control, and evaluate the effectiveness of the control strategy.

Images