You don’t need explicit ratings to do collaborative filtering, by the way. If Ken ordered a movie on Netflix, that means he expects to like it. So the “ratings” can just be ordered/not ordered, and two users are similar if they’ve ordered a lot of the same movies. Even just clicking on something implicitly shows interest in it. Nearest-neighbor works with all of the above. These days all kinds of algorithms are used to recommend items to users, but weighted k-nearest-neighbor was the first widely used one, and it’s still hard to beat.