A Tool for Predicting Music Success

HAMR 2020 Project - https://www.ismir2020.net/hamr/

A python based tool for predicting if a song will appears at Spotify's Top 50 Global ranking based on user inputs.

Requirements

To run this tool you'll need Python 3. Download Python at https://www.python.org/downloads/.

Also, if you do not have some of the python packages needed a pip command will be executed, so I strongly recommend to have pip installed. Installation guide at https://pip.pypa.io/en/stable/installing/.

How to use

First, clone this repository with the git clone command.

To run the tool basically type

python predict.py

If some of the required python packages are missing, it will first install these packages and then run the tool.

When running the tool you can type three commands as follows:

Command	Action
help	Prints helpful information for the user
predict	Predicts if a song will become popular based on user inputs
exit	Closes the program
Other	If anything else is typed, requests a valid option

Links

My master's thesis where I fully explain the model used in the prediction tool (the text is in Portuguese): https://tede.ufam.edu.br/bitstream/tede/7572/6/Disserta%C3%A7%C3%A3o_CarlosAra%C3%BAjo_PPGI.pdf

My ISMIR 2020 LBD Poster were I also present the model methodology: https://program.ismir2020.net/lbd_433.html

HAMR Video Report: https://www.youtube.com/watch?v=nQb-fRebUhk

HAMR slides: https://docs.google.com/presentation/d/1u4AJjzmqWdrmm9-l-d4AN4uQFHU3iyYUrKbT4PmaNPI/edit?usp=sharing

Spotify API documentation: https://developer.spotify.com/documentation/