a new recommendation approach: master's thesis idea

a new recommendation approach: master's thesis idea

For my graduation project, I have a music recommendation idea to realize in 4-5 months. This project is funded by the school, too. The program I'm enrolled in is a mix of music and technology. That means at some point I need software developers which is the deadliest part since I am only a musician.


the idea is a combination of music genre retrieval, analysis, and recommendation monitoring, at its simplest form. I am aiming to create something that can work simultaneously with Spotify. This includes using the audio data of Spotify, too. As far as I understand Spotify does provide such opportunities in its API kit. 


The fundamental algorithm qualities I aim for are given right below: (The secret sauce is not mentioned there)


  1. Track down user’s listening behavior or use existing input derived from Spotify(If possible)
  2. Analyze metadata (genre-focused)
  3. Filter Genres (according to the input has been made over time)
  4. Group those tracks
  5. Serve them as separate playlist within Spotify

finding the guy is the main issue here! First of all, I need to know what type of developer can handle it, and what my limits are when it comes to acquiring/editing/reshaping information from Spotify.



1 Reply

These are logically chosen qualities of the algorithm, I think you had no problem with that. What are your results if you continued this work?
To be honest, I find such tasks as a master thesis very difficult for me and I would rather turn to writing services for help right away. For example, at you can learn about the writing service PapersOwl, which can help you quickly get the coveted quality assignment material. I think that such services are acceptable for many students, think about it for the future.

Suggested posts