AI-powered personalized advertising system

Abstract

An AI-driven system for personalized advertising in real time, where: ◯ an analytics unit to monitor and analyze user behavior in real time across multiple digital platforms such as websites, mobile applications, social media, and smart devices; the unit collects data on user engagement, browsing patterns, time spent, and content preferences to enable targeted, personalized advertising; ◯ a prediction module that predicts preferences and interests of a user, operatively connected to the user interaction data collection unit, wherein the prediction module uses machine learning models such as deep learning, recurrent neural networks (RNNs) and transformer-based architectures to predict interests of users based on historical interactions and inferred preferences; ◯ an emotion and sentiment analysis unit that assesses the user's mood in real time through computer vision, natural language processing (NLP) and voice analysis, whereby the analysis of facial expressions, voice pitch and linguistic mood is used to determine emotional states and receptivity to advertising content; ◯ an embodiment of a context awareness component in operational communication with the emotion analysis and mood unit, in some cases further augmented by various environmental and situational data such as the device type and its physical location, date and time, and the content processed in the device, processing methods, etc., in establishing adaptability and automatic ad placement to be as relevant as possible to the user and their status as prescribed; ◯ Use reinforcement learning algorithms and generative AI models to drive advertising with personalization engines. Creative elements, messaging, and presentations are dynamically adjusted in real time based on user responses to ensure advertising is personalized and always optimized for best performance; o a privacy-focused AI system with federated learning, differential privacy methods, and on-device AI processing to reduce targeted advertising while complying with international data protection laws such as the General Data Protection Regulation (GDPR) and California consumer privacy laws; o an operationally adapted ad delivery mechanism to engage with real-time bidding (RTB) networks, programmatic advertising exchanges and demand-side platforms (DSPs) and place advertisements through digital advertising networks, which guarantees the delivery of tailored advertising to the most relevant audience in real time; and o a contextual feedback loop in which the machine learning models used in the preference and interest prediction module and in advertising personalization The engine is continuously updated to reflect the latest user engagement data, improving personalization over time and optimizing advertising performance.

Description

Field of the invention:

[0001] The invention relates to a system for real-time personalization of advertising presented through an artificially intelligent video advertisement based on viewer behavior, impressions, and patterns. The novelty of this invention lies in the use of machine learning algorithms, deep learning, and sentiment analysis methods applied to a user's digital footprint, which is derived from their web history, content interaction, and purchase data, as well as biometric feedback—i.e., facial expression, voice tone, and, where appropriate, physical response. Through advanced natural language processing (NLP), computer vision, and reinforcement learning, the system dynamically modifies the advertisement to ensure maximum relevance, engagement, and conversion rates, while seamlessly and unobtrusively complementing the user experience. Background of the invention:

[0002] With the rapid development of digital marketing and advertising technologies (AdTech), the focus is increasingly shifting towards personalized advertising to improve user interaction and conversion rates. In the past, digital advertising heavily relied on rule-based targeting and static segmentation approaches such as demographic profiling, keyword targeting, and cookie-based tracking. Although these approaches have contributed to better targeting of advertising campaigns over time, limitations still exist, such as over-targeting, ad fatigue, irrelevant content, and low user interaction.

[0003] With the advancement of artificial intelligence (AI) and machine learning (ML), advertisers are increasingly turning to predictive analytics and automated optimization methods to create more tailored ads. Most of today's advertising systems rely heavily on past behavioral data and explicit user interactions, such as browsing history, search queries, and past purchases. These methods may miss temporary behavioral shifts, fleeting interests, and spontaneous emotional reactions, which can lead to poorly targeted ads and missed engagement opportunities.

[0004] In addition, there are increasing regulations such as the European General Data Protection Regulation (GDPR) and the California Consumer Privacy Act (CCPA) in the US, which severely restrict the collection and processing of user data. The end of third-party cookies in mainstream web browsers and the growing need for data protection as a top priority have increased complexity and forced companies to adopt compliant and privacy-friendly personalization techniques.

[0005] Given these challenges, the industry needs an AI-driven, personalized advertising system that flexibly adapts to each user's browsing and emotional responses while ensuring data privacy and regulatory compliance. The present invention is an AI-based advertising generation system that can tailor advertising across all digital platforms—web, mobile apps, social networks, and connected smart devices. This control system is based on a combination of: 1. User behavior analysis: The ability to track engagement signals, such as what users scroll for or how much time they spend in certain sections, what they do or consume in real time, and infer their intent. 2. Preference learning: We applied deep learning models to continuously learn user preferences in the given context as users provide more feedback and new interests are discovered. 3. Emotional and sentiment recognition uses computer vision, facial expression analysis, voice recognition and sentiment analysis to measure a user's emotional state and assess receptivity to the advertisement. 4. Context awareness: Evaluating live environmental elements such as device type, geographic location, time of day, ongoing activity, and content context to enable effective ad placement. 5. Adaptive real-time personalization by building reinforcement learning and generative AI models that constantly and automatically change ad creatives, messages, and timing in response to live user interactions. 6. AI techniques for privacy protection, integrating federated learning, device processing, and differential privacy in approaches to avoid directly collecting data and comply with global data protection regulations such as the GDPR.

[0006] By leveraging these features, the platform also enables advertisers to create hyper-personalized and emotionally resonant ads, helping companies increase user engagement and maximize marketing ROI. Compared to traditional static advertising models, this invention intelligently varies the type of ads in real time based on subtle user granularity. It learns user activity in relation to various factors in real time, thus increasing overall advertising effectiveness.

[0007] The system would also be cross-platform and could be implemented on web portals, mobile applications, connected TVs, smart speakers, and IoT-enabled devices. This ensures that the advertising experience remains personalized and consistent across different devices and platforms—from one broadcaster to another. This innovation represents a leap forward in the world of AI-powered digital advertising, addressing the weaknesses of existing technologies and providing a privacy-conscious, intelligent, and personalized advertising experience. Summary of the invention

[0008] The system uses real-time data processing and machine learning to create targeted ads that appeal to individual customers while keeping pace with constantly changing market trends and consumer behavior. Traditional advertising methods rely on predefined segmentation and historical data tracking, resulting in irrelevant ad placements, low engagement, and ad fatigue. This invention leverages cutting-edge AI methods such as machine learning, deep learning, reinforcement learning, and multimodal datasets to address these shortcomings by dynamically optimizing ad delivery based on real-time user engagement and predicted interests.

[0009] To summarize its principles: The system combines multiple streams of real-time data to create a hyper-personalized and adaptive advertising experience. Real-time activity tracking: This component tracks user behavior data during their interactions in real time to tailor the experience based on inferred user intent and ad acceptance, such as scrolling behavior, time on page, click-through rates, usage patterns, and browsing history. While traditional methods use a static behavior model, this system adapts on the fly, so ad content is tailored to changing user activity. Preference and interest prediction module: This module uses predictive analytics and deep learning models to examine historical engagement trends. This enables the system to predict user preferences and tailor advertising accordingly.This system can analyze long-term behavioral trends and provide users with appropriate advertising by using recurrent neural networks (RNNs) and transformer-based models.

[0010] A key innovation of this system is emotion and sentiment analysis. The system detects signals of human emotional states and responds using computer vision, natural language processing (NLP), and speech analysis. The system reads facial expressions, eye movements, tone of voice, and the sentiment of the language used in context to assess a user's emotional state and their potential receptivity to advertising content. Regardless of distance, we ensure that ads are presented at the right moment and at the right time. The contextual awareness and adaptive delivery component considers external environmental conditions such as device type, location, time of day, current activity, and digital environment, adapting the ad type, message, and placement accordingly.A user watching a video on a smart TV might see an unobtrusive overlay ad in a split second, while a user scrolling through their feed on a smartphone might see a mobile-optimized interactive carousel ad that they are most likely to interact with at that moment.

[0011] While most advertising systems rely on predefined ad creatives, this invention uses reinforcement learning and generative AI models to dynamically adapt ad content based on live feedback. Based on a user's preferences and responses, the system can adjust ad text and imagery, colors, call-to-action placement, and audio components to ensure each ad is optimized for engagement and conversion. Furthermore, the invention incorporates privacy-preserving AI technologies. Given the growing global competition for user data and privacy, this system leverages federated learning, on-device computing, and differential privacy. The system complies with data privacy regulations (e.g., GDPR, CCPA) by performing ad personalization computations locally on the user's device, reducing the need for centralized collection of user data.This enables hyper-personalized advertising without privacy concerns.

[0012] To be universally applicable, this system is designed for web apps, mobile apps, social networks, streaming services, connected TVs, smart speakers, and IoT devices. This cross-device functionality means that advertising is personalized and optimized throughout the entire customer journey and on every device or platform the customer uses. The architecture generates more relevant and emotionally connected, natural, and frictionless advertising, and ensures significantly less consumer-intensive advertising by combining AI-based automation, real-time learning and adaptation, and privacy-focused technologies.

[0013] This breakthrough enables a completely new form of digital advertising, powered by intelligent, contextual, and privacy-focused AI, delivering immersive experiences to consumers worldwide. This invention helps publishers improve their perceptions by integrating user sentiment, SCR, and disclosure. Advertisers benefit from hyper-targeting users with context-aware advertising brand products. Marketers also gain deep insights into user behavior—the benefits are enormous. Users can seamlessly deliver ads across any device and platform, while dynamically adapting ad creatives.

[0014] At the core of this invention is a highly efficient, automated, intelligent, and real-time personalized advertising system powered by artificial intelligence (AI). It promotes interaction between visitors and advertisers, increases brand and consumer engagement, and boosts advertising ROI. Deep learning, sentiment analysis, contextual adaptation, and privacy-preserving AI – the combination of these evolutionary elements makes this system an innovative approach to intelligent, customer-centric, and compliant digital marketing. Description of the invention

[0015] Fig. shows a block diagram of an AI-driven personalized advertising system. This system 100 includes a dynamic advertising personalization framework with several integrated components for user interaction, relevance, privacy compliance, and advertising efficiency.

[0016] A user interaction data acquisition unit 102 monitors and analyzes real-time user interactions across various digital platforms, including websites, mobile applications, social networks, and smart connected devices. Another unit captures behavioral data such as click rates, dwell time, content read, browsing history, and interaction patterns, enabling a constant data stream for AI-driven advertising personalization. The module under consideration is linked to the field-programmable AI processing unit 102(a), which implements adaptive neural network models capable of processing data acquired during a high number of user interactions in real time.Our AI processing units are flexible and can be reconfigured on the fly, allowing our models to adopt the personalized learning strategies that work best based on the unique characteristics of each ad.

[0017] The preference and interest prediction module 104 is operatively coupled to the user interaction data acquisition unit and leverages state-of-the-art deep learning models, including recurrent neural networks (RNNs) and transformer-based networks, to dynamically predict user interest. This module analyzes historical interactions, evolving user preferences, and inferred user intent in real time to optimize advertising targeting. This allows the system to accurately assess the evolution of user engagement and dynamically adapt its advertising approaches.

[0018] An automatic emotion and sentiment analysis unit 106 uses computer vision, symbolic natural language processing (NLP), and voice recognition to analyze facial expression, vocal intonation, and grammatical sentiment. This allows the advertisement to dynamically adapt in real time to the user's emotional state, cognitive engagement, and receptivity to different content types. The system also uses biometric data analysis, where applicable and permitted by law, to better understand the audience's emotions, adding an additional layer of personalization to the advertisement.

[0019] A context awareness module 108 also analyzes environmental and situational variables such as device type, location, time of day, current activity, and the context of content usage in real time. This ensures that ads remain relevant to the user's immediate digital context, thereby increasing ad relevance and reducing intrusiveness. This module focuses on delivering ads in the right context by leveraging geospatial analytics and real-time event tracking to significantly improve the user experience.

[0020] The system includes a decision-making unit known as the AI-powered advertising personalization engine 110. The modified algorithm relies on reinforcement learning techniques, generative AI models, and other multimodal processing methods to adapt ad creatives, messaging, visuals, and calls to action in real time. This differs from static ad targeting systems, where the engine learns in real time from actual user responses, thus improving advertising effectiveness and engagement. The AI ​​engine does this autonomously, adapting ads in real time to user behavior and feedback, thus achieving high engagement rates.

[0021] The privacy-preserving AI system 112 is a key element of the system and complies with global data protection laws, including GDPR and CCPA. This includes federated learning and on-device AI processing. The system leverages federated learning, on-device AI processing, and differential privacy to achieve personalization without compromising security or collecting personal data.

[0022] The advertising delivery system 114 distributes, delivers, and optimizes personalized advertising across various digital advertising platforms (e.g., real-time bidding (RTB) networks, programmatic advertising systems, social media advertising, search engine marketing (SEM), etc.). This module interacts directly with advertising exchanges and DSPs to ensure ad placement is optimized based on AI-based predictions and insights into user behavior. This ensures advertising consistency across a user's various individual touchpoints, including cross-device synchronization.

[0023] The feedback classifier is rounded out by the real-time feedback loop 116. This receives recommendations from the ad targeting system and optimizes parameters in real time by dynamically modifying AI models and improving recommendations for future interactions. This self-adaptive framework allows ads to self-improve over time based on actual user feedback, resulting in a highly responsive and adaptive advertising ecosystem.

[0024] The user interaction data collection unit 102 collects user interaction data using information from various ecosystem platforms and integrates privacy-preserving analytics to minimize user tracking that links advertising to privacy concerns. The product and interest prediction module 104 leverages a hierarchical, feature-based deep learning model to correlate historical interactions with behavioral patterns, thus providing highly accurate, real-time predictions of interests that match an advertisement.

[0025] The emotion and sentiment analysis unit 106 combines real-time facial recognition, real-time sentiment token extraction using NLP, and biometric analysis to create emotionally relevant advertising systems that adapt to the user's current emotional / cognitive state. Data for the context awareness module 108 comes from geolocation analyses, device-based usage patterns, and content-based behavioral analysis processes, demonstrating that advertisements are almost always displayed in a targeted manner in the digital space.

[0026] Another aspect of this invention, the AI-powered advertising personalization engine 110, utilizes multi-agent reinforcement learning algorithms to test and optimize various advertising variations, achieving optimal results in real time. The AI ​​system protects user privacy through differential privacy and homomorphic encryption, enabling personalized advertising without compromising user data. The federated learning methods can continuously improve the AI ​​models without exposing raw user data, thus addressing modern privacy concerns.

[0027] Data Encryption and Compliance: To ensure secure and compliant ad delivery in the AI-based ecosystem, the system relies on end-to-end encryption for user data transactions, blockchain-based publication reliability tracking for ad content, and a zero-trust architecture for ad delivery networks. Homomorphic encryption techniques ensure data confidentiality while safeguarding the advertising effectiveness of the data processed in the system.

[0028] Furthermore, the system complies with global data protection standards and integrates consent-based advertising structures that give users complete control over their personal data and ad recommendations. It's a next-generation, AI-based, real-time adaptive advertising engine that delivers the right ad at the right time. Advanced machine learning techniques enable emotion-based ad targeting tailored to the user's current emotional state, while contextual awareness dynamically selects relevant ads that align with the user's environment.Leveraging cutting-edge reinforcement learning and machine learning techniques, this solution not only creates a highly personalized and effective advertising experience, but also respects users' notification preferences and companies' collection of user interaction data. It represents a key innovation in data-driven advertising.

[0029] This block diagram ( Fig. ) illustrates at a high level the modular architecture of the system and its seamlessly integrated components for AI-driven personalization, real-time emotional and behavioral analytics, and privacy-focused ad delivery mechanisms for optimal engagement and marketing performance.

Claims

[1] An AI-driven system for personalized advertising in real time, where: ◯ an analytics unit to monitor and analyze user behavior in real time across multiple digital platforms such as websites, mobile applications, social media, and smart devices; the unit collects data on user engagement, browsing patterns, time spent, and content preferences to enable targeted, personalized advertising; ◯ a prediction module that predicts preferences and interests of a user, operatively connected to the user interaction data collection unit, wherein the prediction module uses machine learning models such as deep learning, recurrent neural networks (RNNs) and transformer-based architectures to predict interests of users based on historical interactions and inferred preferences; ◯ an emotion and sentiment analysis unit that assesses the user's mood in real time through computer vision, natural language processing (NLP) and voice analysis, whereby the analysis of facial expressions, voice pitch and linguistic mood is used to determine emotional states and receptivity to advertising content; ◯ an embodiment of a context awareness component in operational communication with the emotion analysis and mood unit, in some cases further augmented by various environmental and situational data such as the device type and its physical location, date and time, and the content processed in the device, processing methods, etc., in establishing adaptability and automatic ad placement to be as relevant as possible to the user and their status as prescribed; ◯ Use reinforcement learning algorithms and generative AI models to drive advertising with personalization engines. Creative elements, messaging, and presentations are dynamically adjusted in real time based on user responses to ensure advertising is personalized and always optimized for best performance; o a privacy-focused AI system with federated learning, differential privacy methods, and on-device AI processing to reduce targeted advertising while complying with international data protection laws such as the General Data Protection Regulation (GDPR) and California consumer privacy laws; o an operationally adapted ad delivery mechanism to engage with real-time bidding (RTB) networks, programmatic advertising exchanges and demand-side platforms (DSPs) and place advertisements through digital advertising networks, which guarantees the delivery of tailored advertising to the most relevant audience in real time; and o a contextual feedback loop in which the machine learning models used in the preference and interest prediction module and in advertising personalization The engine is continuously updated to reflect the latest user engagement data, improving personalization over time and optimizing advertising performance. [2] The system of claim 1, wherein the user interaction data collection unit is adapted to aggregate cross-platform user interaction data using a minimum of tracking via privacy-preserving analytics, thus enabling personalized advertising without compromising user privacy. [3] The system of claim 1, wherein the preference and interest prediction module uses a multi-stage deep learning approach, in which the first stage identifies potential areas of interest of the user based on historical activity trends and the subsequent stage uses reinforcement learning to optimize the accuracy of the prediction in response to immediate user reactions. [4] The system of claim 1, wherein the emotion and mood analysis unit comprises a biometric signal processing sub-module capable of analyzing biometric data with high temporal resolution, such as heart rate variability, pupil dilation, and micro-expressions, to improve the accuracy of emotional state assessment and thus enable even better prediction of advertising receptivity. [5] The system of claim 1, wherein the context awareness module further comprises real-time environmental tracking that integrates geolocation services, user device behavior, and browser behavior analytics to dynamically tailor advertising based on the user's immediate online and offline environment. [6] The system of claim 1, wherein the AI-powered ad personalization engine is further trained to determine the execution of AI models based on the complexity of user engagement in consuming the advertisement, wherein advertisements requiring higher personalization complexity implement advanced deep learning architectures, while advertisements with lower complexity use lightweight AI models, thereby maximizing computation and reducing latency. [7] The system of claim 1, wherein the privacy-preserving AI system further comprises a secure multi-party computation module in which multiple advertisers and multiple AI models train on shared datasets without directly exchanging user data, so that the accuracy of ad targeting is further improved while protecting users' privacy indicators. [8] The system is described in claim 1, wherein the advertising delivery system comprises an integrated fraud detection system that utilizes artificial intelligence (AI) algorithms that analyze data in real time to identify and exclude fraudulent clicks, bots, or non-human interactions. This ultimately improves the return on investment (ROI) of advertising and ensures that ads reach genuine users. [9] The system of claim 1, wherein the real-time feedback loop includes a reinforcement learning module that continuously refines the advertising targeting tactics, allowing the AI ​​frameworks to evolve according to changes in user preferences, seasonal fluctuations, and new engagement metrics. [10] The system of claim 1, wherein the advertising delivery system is further adapted to contextualize third-party behavioral databases and AI models to include advertisements, and wherein the system further regularly collects advertising interaction data from network-based cloud advertising intelligence services to update machine learning models in real time, thus leveraging the constantly evolving user or advertising interaction data to increase the behavioral targeting factor of the advertisement.

Images