merytouceda/csss-spotify

Home of the Spotify data analyses performed by the CSSS'22 student group.

Spotify Analysis CSSS 2022

Home of the Spotify data analyses performed by the CSSS'22 student group.

Questions

Zach:

• Complexity against listening counts and comparison between genres (Complexity metrics: language complexity, compressed file size)

Ben:

• Changes in properties in time

Elise:

• Topics in top songs across ages.
• Ask people in CSSS to download their top 50, and analyze differences.

Mery:

• Diversity of spotify made playlists vs. user made playlists. Special focus on "PRIDE" playlists.
• Differences in topics between pre-post covid breakout released songs. Perhaps sentiment analysis.

Simone:

• Evolution of clulture from music data

Lena:

• Recommended vs. music selection.

Tools

• Scattertext: https://github.com/JasonKessler/scattertext/blob/master/README.md

Resources

• (O'Toole & Agnes Horv´at)[https://arxiv.org/pdf/2206.07754.pdf]. Novelty and Cultural Evolution in Modern Popular Music

CSSS admin

Describe briefly your idea We are using spotify data to explore cultural trends, in the CSSS cohort, and also in the general population. The focus is to learn and create cool data visualizations and maybe make some cool findings on the way.

Do you have a research question? (It's ok if the answer is "not yet") (See above)

What is everyone's role in the group? Ben: API coding and visualization help. 90-10, CSSS-after. Simone: passive learning, some questions. Zach: coding and viz. No data collection. 90-10, CSSS-after. Elise: coding, data analysis, personal questions. Summer school data collection. Will to continue after. Lena: peripheral, interested in CSSS collection. Mery: admin guru + leading.

What do you expect to achieve by the end of summer school?

1. Data collection on CSSS cohort
2. Visualization of a couple of smaller projects.

Data collection

Message we can put on discord
Hi everyone,

For the Spotify project, we would like to analyse music preferences of CSSS participants. If you would like to help us include as many people as possible in our analysis and don’t mind sharing some of your listening history with us, we would be super grateful 🙂

The data we are interested in for our analysis is on users’ top items (artists and tracks) and their Discover Weekly playlists. If you are happy to give us permission to access your Spotify data, we will only be using this data.

The way it works: Spotify has an API through which users can extract their own listening history, but Ben has created a web app, through which it is super easy for you to give us permission to access your account’s data. Here is the link for it: example.com You can of course easily revoke the access at any time by …

Thank you so much in advance!