This paper surveys existing works on human-in-the-loop from a data perspective and classify them into three categories with a progressive relationship: the work of improving model performance from data processing, the work through interventional model training, and the design of the system independent human- in- the-loop.