Please follow the student version of the notebook, fill the gaps and have fun doing so. The very open exploratory questions at the end are just some suggestions for further exploration of the dataset, feel free to follow your instincts and come up with any other interesting things to try.
The solution can be previewed on nbviewer
The exercise is based on the MovieLens 1M dataset. You can manually download it using this link and unzip it on the same folder where the Notebook is running or just run
bash setup.sh