Machine Learning Recaching for Dynamic Page Content

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current caching techniques for dynamic and complex websites lack real-time decision-making capabilities, leading to suboptimal caching strategies and potential user experience issues due to outdated cached content.

Innovation Solution

A system and method utilizing Machine Learning (ML) models to dynamically assess the worthiness of updating cached pages by generating scores based on attributes and network parameters, determining recaching actions for pages.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If traditional caching techniques are used to store copies of pages, then server load is reduced and load time is improved, but content freshness deteriorates due to lack of real-time update decisions

Engineering Contradiction:

Improveload timeVSAvoidcontent freshness

Core Design Contradiction:

ProductivityVSLoss of information

Solution Approach 1:

The patent applies dynamics by transitioning from static caching strategies to dynamic recaching decisions driven by real-time ML model predictions. The system continuously evaluates changing conditions (user behavior, content characteristics, network state) to determine optimal recaching timing, making the caching strategy adaptive rather than fixed.

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The patent implements feedback through ML models that analyze real-time data about page access patterns, content changes, and network conditions. This feedback loop enables the system to learn from past performance and continuously optimize recaching decisions, balancing content freshness against resource efficiency based on actual system state.

Inventive Principle:

Principle #23Feedback

2Loss of information

If real-time recaching decisions are implemented using ML models, then content freshness is improved, but device complexity increases

Engineering Contradiction:

Improvecontent freshnessVSAvoidsystem complexity

Core Design Contradiction:

Loss of informationVSDevice complexity

Solution Approach 1:

The patent introduces an intermediary ML model layer between the content delivery system and the caching mechanism. This intermediary processes complex real-time data and translates it into simple recaching decisions, shielding the rest of the system from complexity while enabling intelligent, real-time content freshness management.

Inventive Principle:

Principle #24Intermediary (Mediator)

3Reliability

If frequent recaching is performed to maintain content freshness, then user experience is improved, but resource efficiency deteriorates

Engineering Contradiction:

Improveuser experienceVSAvoidresource efficiency

Core Design Contradiction:

ReliabilityVSLoss of energy

Solution Approach 1:

The patent changes the parameter of recaching frequency from fixed to variable, controlled by ML model predictions. The system adjusts recaching intervals based on predicted content changes, access patterns, and network conditions, performing recaching only when parameters indicate it will provide value, thus balancing user experience against resource consumption.

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS20250307333A1Method for managing updates to cached pages and system thereof

Publication Date: 2025.10.02 INFOSYS LTD

AI summary

The disclosure relates to method and system for managing recaching of pages. The method includes extracting a set of attributes associated with a page. The method includes generating a set of first scores and a second score. Each of the set of first scores is generated based on an associated subset of the set of attributes and the second score is generated based on a set of network parameters. The method includes determining a recaching action for the page, based on the set of first scores and the second score using a Machine Learning (ML) model.