Lyrics Analyzer
-
NLP
- Download any set of lyrics/discography from one artist (Kanye: https://www.kaggle.com/slickwilly/kanye-west-discography)
- EDA
- Create a list of questions to answer?
- What are the most used words? Themes?
- Artists favourite car? Designer brand? Food?
- Number of unique words? Vocabulary?
- Similies? Like/as usage
- Length? Number of words in the track? Normalize over time? (words/min)?
- Create a list of questions to answer?
- Predict and validate
- Create a song?
- Predict what theme is based off verses/choruses?
-
Application
- Made a web(?) app to retrieve artist data in real time
- Store data/lyrics
- Either web scrape or rate limit on genius api (Python https://github.com/johnwmillr/LyricsGenius Generalhttps://docs.genius.com/)
References: General Sentiment Analysis https://medium.com/@kellylougheed/coding-english-lit-natural-language-processing-in-python-ba8ebae4dde3 Sentiment Analysis for Lyrics https://kvsingh.github.io/lyrics-sentiment-analysis.html