Definitely some clustering method based on similarity of the vectors (there are many, pick a simple one to start)