Mood-based playlist recommender for Indian Classical Music
This project can be partitioned in four parts: (1) database creation (2) exploratory data analysis (3) developing models (4) validating models
The database is created on PostgreSQL and the code used for this are separate .py files, one for each functionality. To know more, read README_DB to understand the flow.
If you are interested in data science, then you can ignore the databse creation part. For EDA, the IPython notebooks "Exploratory Data Analyis" and "Understanding Bias" would suffice. Models are developed and analyzed in IPython notebook "Unsupervised Clustering." Model is validated in Ipython notebook "Validation" along with two .py files that encode survey generated data.
To know more about the project, read here. To use the App, go here. Note that you would need a Spotify account for using the App.