Generative models and NLP in course mining
Generative models (Generative Adversarial Networks, GAN ) allow us to automatically decompose course content into generative entities providing a base for knowledge maps and student recommendations.
Example:
Using StyleGAN (Python framework ) on a set of course question could provide further insight into the course structure.
Reference:
GIT repository https://github.com/NVlabs/stylegan on StyleGAN.
Author of the tip:
Andrej Košir
University of Ljubljana