This thesis focuses on solving the $K$-means clustering problem approximately with side information provided by crowdsourcing. Both binary same-cluster oracle and general crowdsourcing framework are considered. It can be ...