Which of the following statements about Document clustering is true?

(A). In this context, a cluster center in K-Means is represented by the Median of the objects within a cluster.

(B). Expectation Maximization fall within a subcategory of the flat clustering algorithms, called Model-based clustering. 

(C). Agglomerative methods start with an initial clustering of the term space, where all documents are considered to represent a separate cluster

(D). Divisive clustering algorithms start with a single cluster containing all documents.