music-inference

Using an OWL Ontology and neural network to classify songs and recreate them in different genres.

Setup:

Install Necessary Dependencies

owlready2: OWL python API

• documentation and project site: https://pypi.org/project/Owlready2/
• pip3 install Owlready2

music21: MIDI and musicology libraries

• documentation and project site: http://web.mit.edu/music21/doc/index.html
• pip3 install music21
• dependency: matplotlib
• pip3 install matplotlib

BeautifulSoup: web scraping

• documentation and project site: https://www.crummy.com/software/BeautifulSoup/bs4/doc/
• pip3 install beautifulsoup4

pandas: data analysis & manipulation

• documentation and project site: http://pandas.pydata.org/pandas-docs/stable/
• pip3 install pandas

requests: http library

• documentation and project site: https://2.python-requests.org/en/master/
• pip3 install requests

tensorflow: machine learning library

• documentation and project site: https://www.tensorflow.org/
• pip3 install tensorflow

keras: wrapper API for tensorflow

• documentation and project site: https://keras.io/
• pip3 install keras

Import libraries as needed in a python script

• import owlready2 as owl
• import music21 as music
• from bs4 import BeautifulSoup
• import pandas as pd
• import requests
• from keras import Sequential

TODO:

• Figure out inputs, layers, outputs, etc.
• Train our network.

Resources and Examples:

• MIDI instruments and their associated roles