Detoxify: An automated system to recalibrate YouTube recommendation algorithms using intent based content surfing

YouTube's recommendation system relies heavily on what users watch to decide which videos to suggest next. While this usually helps personalize content, it can quickly go wrong. Even watching one unrelated or misleading video can confuse the algorithm and fill the feed with irrelevant suggestions. As a result, users often end up spending time on content they don't actually care about. To tackle this issue, this project introduces Detoxify, a Chrome browser extension designed to help users regain control over their YouTube recommendations by automatically retraining the algorithm back to what truly interests them. Detoxify works by automatically opening curated, topic-specific YouTube videos in background tabs. After the user selects a topic, for example, Rust programming, machine learning, or fitness, the extension searches YouTube Data API for relevant long-form videos, opens them in inactive tabs for a configurable duration (default 30 seconds each), and closes them automatically. Over time, this background engagement process helps YouTube's algorithm relearn the user's real interests and clear out unwanted recommendations. The main goal of Detoxify is to promote a healthier and more focused online experience. By reducing exposure to distracting or low-value content, it helps users keep their recommendations meaningful and relevant. The project also considers the ethical and technical aspects of automating such interactions, ensuring the system operates responsibly, entirely client-side, with all user data stored locally and no server communication. In essence, Detoxify is a simple yet powerful solution that helps users reset their YouTube experience and make the platform work for them again. The Detoxify project successfully addresses one of the most common issues faced by YouTube users: the distortion of personalized recommendations caused by unintended viewing behavior. By automating the process of opening curated, high-quality videos aligned with the user's chosen interests, Detoxify effectively retrains YouTube's recommendation algorithm to reflect genuine user preferences. The system's design focuses on user privacy, minimal resource consumption, and compliance with ethical standards, ensuring that users regain control of their digital experience without the need for manual intervention or server-based infrastructure. Through its smart automation, curated content selection, and background execution features, Detoxify enhances digital well-being by minimizing exposure to distracting or irrelevant content. It not only improves the relevance of YouTube recommendations but also contributes to a healthier, more productive online environment. The project demonstrates how lightweight, ethical automation can positively influence algorithm-driven ecosystems when used responsibly. By operating entirely client-side with all data stored locally, Detoxify respects user privacy while delivering measurable improvements in recommendation quality. The Algorithm Shift Score provides transparent feedback on progress, motivating users to engage in ongoing algorithmic retraining. Overall, Detoxify represents a practical, user-centric solution to a widespread problem in modern digital platforms.