Groups of similar documents can be formed, and are called clusters.

Each cluster can be represented by its average, or centroid. See the "x" in the approximate center of each cluster for the location of its centroid.