I have idea of creating Content based Music recommendation system based on Music signals, Now I am asking how I can extract signal features from Music and how to represent them. There after I may find similarity between them to make recommendation?

Your ideas will help me, Thanks!


It is a very broad question. You need to specify the type of categorization of this recommendation system. Then you can think about which features might be relevant to this system. Generally you need to apply a classification approach on music signals.

| improve this answer | |
  • $\begingroup$ Thank you Remy, the category of this Recommendation is Content Based Recy system. The features I do think to consider are Frequency feature, Signal sequence feature and Temporal features. But how to extract them and representation of them is problem to me?. Thanks $\endgroup$ – Alex M Nov 15 '16 at 6:41
  • $\begingroup$ There are lots of papers and tutorials on music recognition and classification. Please take a look at this: $\endgroup$ – Remy Nov 15 '16 at 20:41
  • $\begingroup$ cms2.unige.ch/fapse/neuroemo/pdf/… $\endgroup$ – Remy Nov 15 '16 at 20:41
  • $\begingroup$ Remy Thanks so much, I am checking through and I have got some other information may be will contribute some... (github.com/tyiannak/pyAudioAnalysis/wiki/3.-Feature-Extraction). Thanks $\endgroup$ – Alex M Nov 17 '16 at 4:47
  • $\begingroup$ Thanks Remy, I have managed to extract some features from music..... now trying to learn how to apply NN for classification ...... $\endgroup$ – Alex M Jan 10 '17 at 7:45

Not the answer you're looking for? Browse other questions tagged or ask your own question.