Joris Meys in Elk: "Wanted: open data suitable for a data sc..."

Wanted: open data suitable for a data science project!

Every year, we ask our students 2nd bach to do a project. We give them a (big) dataset (preferably in JSON or a bunch of files combined), ask some research question, and ask them to perform data cleaning and exploration related to that question using #Rstats.

I've used most obvious choices, so I'm turning to Fedi to find new, interesting datasets. If you have no idea, sharing helps too!

#dataScience #openData #Fedihelp #education

February 23, 2026 at 7:12:32 AM

You could use some animal tracking data, e.g. from movebank. We also have drone images of seals if you are interested.

there's also the LiveMouseTracker datasets, very detailed movement and interaction data of up to 4 lab mice in SQLite format.
See here for the project's website https://micecraft.org/lmt/,
and here for an example dataset: https://micecraft.org/lmt/download/20180110_validation_4_ind_Experiment_6644_e.sqlite

(This project is cool for many other reasons, not least the ikea-style build instructions or the fact that they repurpose Kinect cameras for scientific work as high-res 3D video cameras).

micecraft.org

Live Mouser Tracker - MiceCraft

Replying to

Thanks for the suggestion! I've seen movebank before, but forgot about it. Drone images of seals are absolutely cool, but that's going to be a bit much for an intro course data science I think :-)

The chess server LiChess

releases all chess games played on the server under cc0. 7,507,487,928 games so far, released in monthly batches.

https://database.lichess.org/

database.lichess.org

lichess.org open database

Thanks for sharing, that's a massive collection! Honestly a bit too massive for the task at hand actually, even though I know for a fact some of my students would love this. (they're playing chess during class, the little raskalls)