Yes.. perhaps a captcha at user registration step?
We've actually already implemented this. So this may be an instance in which a person is manually making the accounts, but then uses automation to post spam. (Or perhaps does this manually as well, to an extent).
We can only do so much to block spammers from creating accounts - this part is really hard to distinguish - so most of the preventative work we do is in catching and preventing spammy
posts as they are submitted. Without going into too much detail, our solution for this is a rules set that analyzes the content of the posts to see if it matches certain criteria. This set is regressive, so we can enhance it each time spam is posted with data seeded from what came through, in order to ensure that the same (and broadly similar) spam can't be re-posted. It actually works pretty well - we are positively hammered with spam attempts, and the great majority of them are blocked - but it's not perfect.
The most recent examples users have seen all made it through primarily because they were in Korean, which is not a language that our anti-spam measures have been tuned for (given that it's not what we've encountered in the past). I've just made some changes that I think will help us do a much better job of automatically preventing these Korean posts from appearing, though. We shall see.
My apologies for these posts cluttering up the forum, though. No amount of spam is acceptable to us.