This project involves analyzing about 350,000 song lyrics to understand the frequently used words in different genres of music through ngram models, model the use of certain sentiments over time, analyzing closely associated words through hierarchical clustering, latent sentiment analytics and cosine similarity.