A Context-Aware Intelligent Agent for Adaptive Browser Resource Orchestration

Abstract

Modern web browsers consume a significant amount of system memory when multiple tabs remain open, causing reduced performance, slower response times, and occasional browser crashes. Managing tabs manually is inefficient and interrupts the user’s workflow. This paper presents ALIA (Adaptive Load Intelligent Agent), a Manifest V3 Chrome Extension that intelligently monitors browser tabs and optimizes memory utilization through automated, context-aware tab management. ALIA collects tab status, idle time, pinned state, and audible state through Chrome Extension APIs, applies configurable safety rules to protect active, pinned, audible, and sensitive tabs, and safely suspends eligible inactive tabs using the chrome.tabs.discard() API without closing them. A per-domain adaptive threshold, learned from the user’s own return-visit behaviour, and a dedicated secure placeholder for sensitive sites (banking, healthcare, government, and webmail) allow the system to reduce memory consumption while preserving privacy and user experience. A popup and a docked side-panel dashboard present live statistics on open, idle, and suspended tabs and estimated memory freed. Twelve functional test cases covering the permission model, idle detection, suspension safety rules, adaptive thresholds, sensitive-domain handling, and dashboard synchronization were executed, all of which passed, and their observed on-screen outputs are documented in detail. Planned enhancements include Random Forest for tab-importance classification, Long Short-Term Memory (LSTM) networks for memory-usage forecasting, and Proximal Policy Optimization (PPO) for reinforcement-learning-based suspension decisions, moving the system from rule-based to fully predictive, adaptive browser resource orchestration.