Looks like the vector the authors used was that every registrant can be found at a url such as <a href="http://sxsocial.sxsw.com/users/503" rel="nofollow">http://sxsocial.sxsw.com/users/503</a><p>They just iterated through and scraped what was on each page, I'm guessing. There is no contact address, email address or password info.