In modern world and mature democracies citizens use social media platform’s such as YouTube, Twitter (X), Instagram, Facebook, etc. to express their opinions, views, likes, dislikes about government policies and schemes, political leaders. Many citizens also prefer these social media platforms to register their protest. Sentiment Analysis is one of very effective process to understand citizens opinions, views, likes, dislikes about government policies and schemes, political leaders and alliances. This study uses the VADER sentiment analysis tool, to evaluate and gauge public opinion about two prominent alliances, NDA and INDIA based on YouTube Metadata (tittles, descriptions, likes, views count, comments). TF-IDF a powerful technique used for feature extraction in text analysis. The research combines sentiment analysis with machine learning techniques (Logistic Regression and Random Forest classifiers) to assess and compare public sentiment and political popularity.