Predicting YouTube Fame Through Video Title Analytics

This presentation covers a comprehensive YouTube fame analysis by examining recommendation issues, constructing datasets from 11,000+ videos, and leveraging advanced scraping and metadata extraction tools like Playwright and yt-dlp. Predictive modeling techniques focused on video views, evaluated via MSE, utilized methods including Word2Vec and RoBERTa for enhanced results. It highlights challenges like sample bias and interpretability while providing key insights into fame prediction...