Abstract: Micro-blogging has played an important role in online social networks, and it has become a valuable source of public opinions. So there is a huge challenge to supervise public opinions in micro-blogging network. Incarnated accounts is a special kind of accounts that appeared during the supervision of public opinions. Strengthening the supervision of these accounts has great significance. The precondition for supervision is to discover them. In order to find incarnated accounts, a model targeted their characteristic based on the approach of timing and similarity is proposed, then the effectiveness of the model in Sina micro-blogging is verified.