The Student Room Group

Is my EPQ topic ok?

Hi, I'm planning on doing my EPQ on how Spotify's algorithm/recommendation system works. I'm not dead set on a title right now but I was thinking something along the lines of how has Spotify's algorithm allowed it to succeed in the music industry or how accurate is Spotify's recommendation system. I was also thinking of doing an artefact like programming my own music recommender in Python similar to Spotify and explaining how I made it. I was just worried that the topic was kind of a weird choice but I find it super interesting. Is this like actually doable or should I think of something else??
Reply 1
I'm not doing an EPQ but it sounds good to me. Your EPQ topic would be even more beneficial for you if you're planning on doing programming (or similar) in the future. Ask your EPQ guider/teacher etc. if it would be appropriate. Imo the more 'weird' your topic is the more interesting it is to whoever is marking your EPQ (not saying your topic is weird, what I mean is how specific it is)
It is certainly the start of a good EPQ idea but would not translate into an entire project. When you talk about how accurate it is or how it has led to success for Spotify I feel these are not questions that can really be answered.

Additionally, I am not sure how much information you can actually access about how their algorithm works. It might be better to try and compare the differences in the way that different companies' algorithms work as that could translate into a full project. You can talk about the intended function and how they have developed over time. That might translate into a more straightforward EPQ project that still uses your original idea.
Reply 3
It is good to have an EPQ title that is presented as a question. You also need to be sure that you have enough sources.

https://www.futurelearn.com/courses/epq-success

This is free and good to do.
(edited 1 year ago)
Original post by nedflanders123
It is certainly the start of a good EPQ idea but would not translate into an entire project. When you talk about how accurate it is or how it has led to success for Spotify I feel these are not questions that can really be answered.

Additionally, I am not sure how much information you can actually access about how their algorithm works. It might be better to try and compare the differences in the way that different companies' algorithms work as that could translate into a full project. You can talk about the intended function and how they have developed over time. That might translate into a more straightforward EPQ project that still uses your original idea.


+1 if op could assess accuracy of blackbox better than it's creator, and its shortfalls, you could make a better spotify

you could attack the same question from a more robust standpoint along the lines of where does spotify's algorithm 'fail' but that's difficult to evaluate without knowing what spotify's trying to accomplish

looking at the rise of spotify is interesting, but that might not be from the programming standpoint you're after

Quick Reply

Latest

Trending

Trending