MetaCaptcha uses the proof-of-work (or “client puzzle”) approach for combating spam. Proof-of-work approaches slow down spammers by requiring them to solve computational puzzles before accessing a desired service. The idea is to impose a per-access cost in hopes of reducing abuse. More specifically, the cost to spammers will be computational resources that they devote to solving a puzzle before sending each message. The approach was first outlined by Dwork and Naor. However, Laurie and Clayton demonstrated that a small constant cost would fail to significantly reduce spam and a large one would burden legitimate users. Later, Liu and Camp demonstrated that if puzzle difficulties were based on user reputations then a proof-of-work approach would indeed be successful. Thus, malicious clients (spammers) would need to be issued “harder” (longer to solve) puzzles than non- malicious clients. The new comments on your blog will be sent to our MetaCAPTCHA server for spam classification.