Dynamic Page Prefetch Area Shape Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current computing devices face performance and latency issues when loading network content, particularly on devices with limited resources, due to inefficient prefetching of linked pages, which can burden servers and client devices, and fail to accurately interpret user interactions for optimized prefetching.
Innovation Solution
The technology optimizes page prefetch areas by defining and modifying these areas based on user interaction probabilities, using machine learning to segment users and test different prefetch parameters, allowing for dynamic adaptation to changes in page layout and reducing unnecessary prefetching.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If traditional prefetching is used to load linked pages, then page loading speed may be improved, but server burden and network bandwidth consumption increase significantly
Solution Approach 1:
The system performs preliminary actions by prefetching linked pages before they are actually requested by the user. When a user hovers over or interacts with a link, the system proactively loads the associated page content in the background, so that when the user clicks the link, the page is already prepared and can be displayed immediately, improving perceived loading speed while controlling bandwidth usage through intelligent prediction
Solution Approach 2:
The system changes parameters by dynamically adjusting prefetching behavior based on user interaction patterns, device resources, and network conditions. Machine learning models analyze hovering duration, cursor movement patterns, and historical behavior to predict which links are most likely to be clicked, adjusting the prefetching strategy accordingly to balance speed improvement with bandwidth conservation
2Ease of operation
If prefetching is performed for all linked pages, then user experience may be improved, but client device resources are overwhelmed
Solution Approach 1:
The system applies partial action by selectively prefetching only the most promising linked pages based on predicted user behavior, rather than prefetching all linked pages. By analyzing user interaction patterns and assigning probability scores to different links, the system prefetches only those pages with high likelihood of being accessed, improving user experience while avoiding overwhelming client device resources with unnecessary data
Solution Approach 2:
The system segments the prefetching process by dividing linked pages into different priority categories based on predicted click probability. High-probability links trigger immediate prefetching, medium-probability links may be prefetched with lower priority, and low-probability links are not prefetch ed at all. This segmentation allows the system to manage client device resources efficiently while still providing significant user experience improvements for the most likely navigation paths
3Measurement precision
If prefetching area is expanded to cover more of the page, then more potential user interactions are captured, but false prefetch predictions increase
Solution Approach 1:
The system applies local quality by creating prefetch areas with varying sensitivity thresholds for different regions of the page. Areas around high-value or frequently clicked links have larger prefetch areas with lower thresholds for triggering prefetching, while areas around less important links have smaller prefetch areas with higher thresholds. This localized approach captures more genuine user intent in critical areas while minimizing false prefetch predictions in less important areas, balancing interaction detection accuracy with energy efficiency
Data Source
AI summary
A method for optimizing resource prefetch criteria may include identifying a prefetch criteria for a selectable item, the first prefetch criteria being associated with the selectable item. The first prefetch criteria may be modified to a second prefetch criteria different from the first prefetch criteria, where the modification includes a change of shape of a prefetch area from a first shape to a second shape different from the first shape. The first prefetch criteria may be replaced with the second prefetch criteria.