music recommendation system

Author: Nika Saganelidze
Co-authors: Davit Kokauri, Irakli Tsukhishvili, Nika Saganelidze, Giorgi Saralidze
Keywords: music, recommendation, python, django, recombee, vue, nltk
Annotation:

It is very difficult to find exact information that we need in a huge amount of data on the internet. This includes every kind of media: news, songs, videos etc. This is the problem that recommendation engines try to solve. There can be different variables that these algorithms can take into the consideration and give results according to them. For example, certain media can be fitting for you depending on age group, nationality, gender etc. In our project we try to recommend songs to users matching to their mood and past actions, which can be used to deduce their interests. We also want to give the ‘popularity’ of a music less priority during recommendations, because it results in centralization of views. In a perfect world, our application will be able to recommend music to users which relate to their mood, have high quality and are less popular.