This project was completed in week 4 of the Data Science Bootcamp at Spiced Academy.
The goal is to build a web-scraper to scrape music lyrics, preprocess the obtained text, then predict the artist from some lyrics that a user writes in the command line. For testing, I scraped around 100 song lyrics of Metallica and Iron Maiden from metrolyrics.com and used the texts to train a Multinomial Naive Bayes Classifier, which predicted the band from input lyrics with 68% accuracy.
You can read more details about this project on my blog.
- Clone this repo:
git clone https://github.com/lorenanda/lyrics-classification.git - Install the necessary libraries:
pip install -r requirements.txt - Run
scrape_and_predict.pyand follow the bot's instructions
