Benchmark user-based system.

To compare the performance of item-based prediction we also entered the training ratings set into a collaborative filtering recommendation engine that employs the Pearson nearest neighbor algorithm (user-user). For this purpose we implemented a flexible prediction engine that implements user-based CF algorithms. We tuned the algorithm to use the best published Pearson nearest neighbor algorithm and configured it to deliver the highest quality prediction without concern for performance (i.e., it considered every possible neighbor to form optimal neighborhoods).