Add: New GSoC idea on AI-powered learning recommendations.

[pranavb1234]

Hi @walterbender,

This PR documents the idea we discussed earlier. I’m still thinking through the challenges, and this is an early draft.

I’d really appreciate it if you could review it and let me know what I should change or improve.

Regards,
Pranav Bhatia

[walterbender]

How will we know this is working? Is there some evaluation framework? Is there evidence that the signals you propose to analyze have meaning to the learner?

[pranavb1234]

Hi @walterbender,

Sorry for the delayed reply.
The solution that comes to my mind right now is that there could be two main ways to evaluate whether this is working. First, we can look at behavioral signals, for example, whether learners actually open or act on the suggestions, explore related activities or not. If suggestions are consistently ignored, that would likely mean they are not meaningful or relevant.
Second, we can gather qualitative feedback from teachers and learners to understand whether the suggestions feel helpful and supportive rather than prescriptive.

I’ll continue thinking about better ways to evaluate this and improve the framework.
Regards

[walterbender]

To be honest, it seems a bit backward to build something and then seek feedback rather than seek input (from teachers) to influence your design.

[pranavb1234]

That’s a fair point. I agree it would make more sense to seek teacher input first rather than validate after building.
One possible approach could be to begin with a small teacher consultation to understand what kinds of reflective support would actually be useful, and then shape the system around that.

But I think it may also introduce some challenges. Gathering feedback may slow down development, and insights from a small group of teachers may not generalize across different contexts.There may also be differing expectations among educators, making it difficult to design a solution that works for everyone.

Do you have any ideas in mind that I should explore?