Hello, I'm trying to figure out a plan for improving the existing machine learning backends with GSOC. I've installed the environment and already fixed an easy bug in the tracker so I have a rought understanding of how things work, but I'm having a bit of trouble really understanding how all the parts involved work together. I've read the Analytics API and gone trhgough the related pages in the docs, but without some hands on experience I feel like I'm still falling short of really understanding how the modules work let alone how would a plan for improvement look like.
I think going from examples of functionality to how they work would be a good way of figuring things out. In order to do this I would need a suficiently large dataset to test out existing functions. I could of course generate some minimal data myself, but I suspect for the effort involved that would be a pretty weak return on the time invested. Especially when there are probably numerous datasets that a kind stranger could provide.
Therefore, I have a couple of small requests:
1. Where can I find some good datasets of users interacting with moodle to train some models and observe the ml-backend behavior?
2. Are there any good relatively easy Analytics API bugs that I could tackle to get more familiar with this part of Moodle (I did a quick search, but couldn't really anything relevant)
3. Any other general or specific advice as to how to approach this task of understanding and improving the Insights module is very welcomed