As I continue to make progress on this project, I am excited to see more trends showing up.
This week, I looked at artist longitude versus song popularity as well as tempo vs danceability.
The popular songs were concentrated in North America and India (the two dominant industries of the world), as well as Latin America. This shows that, for example, an artist in Hollywood, CA is more likely to gain popularity off a song rather than someone living in say, Japan.
This week I also looked back at certain things I was unable to accomplish during my project.
- I couldn’t build a futuristic model to predict hit songs –> I simply ran out of time and didn’t have the knowledge.
- I wasn’t able to particularly focus on Bollywood songs due to the lack of data points, however, a broader view of songs did help me gain better understanding of popular songs.
In the upcoming weeks, I will be working on my presentation as well as looking at more features, trying to improve the accuracy of my model.