Pre-fetching Resources via User Segmentation and Edge Caching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In network environments, the speed at which users receive resources from server computer systems to client computer systems is hindered by latency and bandwidth, leading to slower retrieval of large files and negatively impacting user experience, especially when resources are not targeted to the user.
Innovation Solution
Implementing a method to pre-fetch resources by predicting user actions through tracking user features and actions, segmenting users based on common characteristics, and storing pre-fetched resources on a computer-readable storage medium closer to the client system, thereby reducing network hops and improving transmission speed.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If resources are fetched from server computer systems over the network, then users can access stored content, but the retrieval speed is reduced due to network latency and bandwidth limitations
Solution Approach 1:
The system performs preliminary actions by predicting user requests before they occur and pre-fetching the required resources from server computer systems and storing them in intermediate storage locations. This preliminary action eliminates network transmission delays when users actually access the content, as the resources are already positioned closer to the client system.
Solution Approach 2:
The patent introduces intermediary storage systems (such as edge servers or caching mechanisms) positioned between the server computer systems and client computer systems. These intermediaries store predicted resources closer to users, reducing the network path length and transmission time while maintaining efficient resource delivery.
2Productivity
If pre-fetching is performed without accurate prediction, then network operations are wasted, but user experience is not improved
Solution Approach 1:
The system implements feedback mechanisms that monitor actual user behavior and access patterns. This feedback is used to continuously refine and improve the prediction algorithms, ensuring that pre-fetching operations are based on accurate predictions of user needs. The feedback loop prevents wasteful pre-fetching by adjusting predictions based on real user interactions.
Solution Approach 2:
The prediction system is dynamic and adaptive, adjusting its predictions based on changing user behavior patterns, time of day, device type, and other contextual factors. This dynamic approach ensures that pre-fetching operations remain efficient and relevant, avoiding wasteful network operations by adapting to actual user needs rather than using static prediction rules.
Data Source
AI summary
Pre-fetching resources by predicting user actions. Features of multiple users and of multiple client computer systems associated with the multiple users are obtained. From these features, multiple segments are generated, each including a proper subset of the multiple users. Users in the proper subset share a common feature. For a segment, resources that have at least a threshold level of likelihood of being requested by a user included in the segment are pre-fetched, i.e., obtained before the resources have been requested. A client computer system associated with the user in the included segment is identified, and the obtained resources are stored for providing to the user.