rasbt / musicmood Goto Github PK

View Code? Open in Web Editor NEW

411.0 33.0 107.0 54.61 MB

A machine learning approach to classify songs by mood.

License: GNU General Public License v3.0

OpenEdge ABL 57.17% Python 0.16% CSS 0.06% HTML 23.38% Jupyter Notebook 19.23%

song-dataset exploratory-data-analysis lyrics mood machine-learning

musicmood's Introduction

MusicMood

A machine learning approach to classify music by mood based on song lyrics.

This project is about building a music recommendation system for users who want to listen to happy songs. Such a system can not only be used to brighten up one's mood on a rainy weekend; especially in hospitals, other medical clinics, or public locations such as restaurants, the MusicMood classifier could be used to spread positive mood among people.

Dataset Summary

A 10,000-song subset was downloaded from the Million Song Dataset.
Lyrics were automatically downloaded from LyricWikia and all songs for which lyrics have not been available were removed from the dataset.
An English language filter was applied to detect and remove all non-English songs.
The remaining songs were randomly subsampled into a 1000-song training dataset and 200-song validation dataset.

Exploratory Data Analysis

Results

musicmood's People

Contributors

Stargazers

Watchers

Forkers

relationbuilder lgeo3 kirkhadley datahacking tuan3w simplyali241 zhoujialinmumu 466152112 zclfly silky chagge xccds evildevilzhong chengzhizhao fdoperezi eq4 kofie rsethur samsadi aicidnn mrmethodmanman nadia-alam micjagga johnnyeric amirmeisami imclab rohankoid mongonauta s3ni0r shubashree s1gm01d raghavatreya connieshen0503 stingerhouse rshetty22 mrinalarun narashiman terratenney harendranathvegi9 sufuf3 ashish3805 yaozhengjie hamedmp dattatele rhomi lmick002 aryanugroho mahmoudshahen agr-ankush clustersdata rahulrana95 martin72gar yashveersingh18 reloadbrain umraoshweta ddddora pskhatri ma5onic goelaish cristianmtr alabhyavaibhav junthbasnet vnpraveenb afcarl jennychiou machineelearner desireedisco aurelius-ai sunilvarma6369 sprio iriswang1 illseption ekansh97vashistha uvashishtha machine4life iosnitishsharma zestiros anshugoe xmirtunjay harshithags nikitagawde10 sonalthube1004 likitar santhu45482 turjo7 microw 970334034 zaid7860 rickykennedy25 robertdigital bennio maxcodextc gokulsg jonad hadryan tumme ebpasandi codeaudit dincerdogan geeklurnai

musicmood's Issues

Getting started

Hey, I just found this epic looking project on my search for a program that is able to classify a .mp3 (or really any other format)
I want to build a Radio Bot for the Discord platform but the biggest Problem is that the bot way too often chooses a song that is angry/happy after a calm/sad song. How could I get started with training and using this AI or is this project even suited for such task? If it is I would of course give you credit.

Best Regards

how to start the main execution

i dont know where to start from

For-loop to determine the mood for the lyrics text files

I am getting this error message after running

with open (lyrics, 'r') as inputFile:
PermissionError: [Errno 13] Permission denied: 'F:/Music/project/lyrics'

About running the tests and webapp.

I am just starting with Machine Learning, so please bear with me :).

I was just wondering if you could give a list of steps to be done for the tests and acquire the results and also to implement the acquired result to the webapp.

Much appreciated.

How to start with my music

Hey,

I stumbled upon you project and wanted to try it with some of my music.
Can you point me where to start and what to edit?

Thanks alot!

overfitting

Sebastian, isn't your model overfitting as the recall falls sharply when you going from train to test data.

Happy/Sad Rorshack

I love the idea for this. Totally want to try use it to find only happy music, and check that my current playlists are happy. I think, though, that my idea of happy might be different to yours, I wonder how different the results might be with a larger group of human annotaters. I also felt slightly cheated when I saw this, I think a lot of the mood affecting qualities of music are in the actual music, as opposed to the lyrics. For some reason I just assumed you had trained the algorithm on audio files, though now I realise how difficult that will be. I listen to a lot of - to me, happy-sounding - foreign music these days, because I like the music of a human voice but I dont want to be distracted by overthinking the lyrics or tempted to sing along while Im working. Occasionally, Ill google the lyrics, and be horrified by how dark they seem, to me they are a completely different mood.

App code?

Any chance you might add the code for the app itself to this? Couldn't care less what state it's in, but I'm working on a similar thing now (trained model --> web app) and have exactly 0.0% of a clue where to start (or even what to google). A working, deployed example would be absolutely beautiful, even if it's a bit rough around the edges. Thanks again for this, and well done!