User-item combinations that appear in both sets are ignored for the test set, and thus in the evaluation.
The evaluation measures are listed in the ItemPredictionMeasures property.
Additionally, 'num_users' and 'num_items' report the number of users that were used to compute the results
and the number of items that were taken into account.
Literature:
C. Manning, P. Raghavan, H. Schütze: Introduction to Information Retrieval, Cambridge University Press, 2008