The best way to handle abuse is to prevent it from happening in the first place. Here are some actions that, if coming repeatedly and frequently from the same session/user/IP/whatever, should at least raise some red flags:
Registration
Login attempts
Deletion (of anything)
Very many actions that need approval, like by/for/of tagging or group requests
The best way to handle abuse is to prevent it from happening in the first place. Here are some actions that, if coming repeatedly and frequently from the same session/user/IP/whatever, should at least raise some red flags:
Open to further suggestions.
Support fancy phrases like "quadratic falloff".