Computational Advantages

In an IBCF system that pre$-$processes the catalog items to create an item similarity matrix, the computational advantages are that $($more than one option may be correct$)$:

The number of similar items already liked by a user is small

The number of users who have already liked the item is small.

The similar users $($peers$)$ are already stored in the memory, which is an advantage during runtime.

The similar items are already stored in the memory, which is an advantage during runtime.