Video Semantic Clustering with Sparse and Incomplete Tags