I would map all ratings, of all values, to a 1. Practically speaking this is probably more 'accurate'.
Are you using a precision-recall test? they're not terribly informative, though they're about the only thing you can do to evaluate recommendations without ratings. That is, it's testing whether it recommends back already-known items, but, that's not necessarily a good test of whether it's making good recommendations. It could be recommending better stuff and not getting credit. So, i wouldn't necessarily worry.
