Generative AI Content Personalization Using Real-Time Behavior Signals

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing web content environments struggle to provide real-time, user-customized content that reflects individual user preferences and behaviors, leading to suboptimal engagement and satisfaction.

Innovation Solution

A method and system utilizing generative artificial intelligence to collect user content and behavior information, infer preferences, and generate targeted content in real-time, adjusting based on scrolling behavior and content consumption time, while considering provider settings and user consent.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If real-time customized content is provided based on user behavior, then user engagement and satisfaction are improved, but system complexity and processing requirements increase

Engineering Contradiction:

Improveuser customizationVSAvoidsystem complexity

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The system performs preliminary actions by collecting and storing user behavior information (clicks, scrolling,停留 time) as users interact with content. This pre-collected data is then used by the generative AI model to quickly generate customized content without requiring complex real-time analysis during content delivery, thus reducing instantaneous system complexity while maintaining high adaptability

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

A generative AI model serves as an intermediary between raw user behavior data and customized content generation. This intermediary component abstracts the complexity of analyzing user preferences and translating them into personalized content, isolating the complexity within the AI model while keeping the rest of the content delivery system relatively simple

Inventive Principle:

Principle #24Intermediary (Mediator)

2Measurement precision

If generative AI models are used to generate personalized content, then content relevance and user satisfaction increase, but processing time and computational resources increase

Engineering Contradiction:

Improvecontent relevanceVSAvoidprocessing time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

User preference information is inferred and stored in advance based on collected behavior data. When customized content is needed, the system retrieves pre-inferred preferences rather than performing complete preference analysis from scratch, significantly reducing processing time while maintaining high content relevance through the use of pre-processed user insights

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system dynamically adjusts content generation based on real-time user behavior signals such as scrolling speed and停留 time. By detecting when users are engaged or disengaged, the system can dynamically trigger or adjust content generation, optimizing the balance between processing time and content relevance based on actual user needs

Inventive Principle:

Principle #15Dynamics

3Measurement precision

If user behavior information is collected in real-time, then content personalization accuracy improves, but data processing load and privacy concerns increase

Engineering Contradiction:

Improvepreference inference accuracyVSAvoiddata processing load

Core Design Contradiction:

Measurement precisionVSUse of energy by moving object

Solution Approach 1:

The system extracts only the most relevant behavior information (clicks, scrolling patterns,停留 time) needed for preference inference, rather than collecting and processing all possible user data. This selective extraction reduces data processing load and energy consumption while maintaining sufficient accuracy for effective content personalization

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The system uses lightweight, ephemeral data structures and processing methods for handling user behavior information. Behavior data is processed in small batches or individually as events occur, rather than accumulating large datasets for batch processing, reducing instantaneous processing loads and energy requirements while maintaining inference accuracy

Inventive Principle:

Principle #27Cheap short-living objects (Disposable)

Data Source

PatentUS20250384104A1Method and system for providing customized content using generative artificial intelligence

Publication Date: 2025.12.18 PIAMOND CORP

AI summary

A method and system for providing customized content using generative artificial intelligence is disclosed. According to one example embodiment, a method for providing content may include collecting content information of original content and behavior information of a user for the original content, in relation to the original content already provided to the user, inferring preference information of the user based on the content information and the behavior information, generating target content through a generative artificial intelligence model based on the inferred preference information of the user, and providing the generated target content.