Dear INLP students,
some of you asked for some real data on which to practice (with programs, not paper and pencil) the concepts presented in the "Evaluation" lecture.
For this, I think the famous "20 newsgroups" dataset is really appropriate.
It's available in scikit-learn, for instance.
Reza (our assistant) also found an nice blog page, which precisely illustrates evaluation on that dataset and could serve as an inspiring starting point for some practice with this respect: https://krakensystems.co/blog/2018/text-classification.
I hope this addresses your request.
Best,
