Make comparative statistics of datasets

This will be useful for the paper: 

- [ ] average mutual overlap (how many concepts in common)
- [ ] size per dataset
- [ ] columns per dataset
- [ ] what type in dataset (norms, ratings, relations)

etc.