Hi all,
I've started a directory for the various scoring modules we might want to play with. The first one I've added is a simple script which calculates the Dice coefficient between a raw query and a URL (basically, character overlap). This should ideally be combined with the existing distributional semantics scoring system. But we can talk about this once your new code is ready.
In case I've done anything wrong with GitHub, please advise!! :-)
Cheers.
Hi all, I've started a directory for the various scoring modules we might want to play with. The first one I've added is a simple script which calculates the Dice coefficient between a raw query and a URL (basically, character overlap). This should ideally be combined with the existing distributional semantics scoring system. But we can talk about this once your new code is ready.
In case I've done anything wrong with GitHub, please advise!! :-) Cheers.