Sorry if I didn't make that clear: I wasn't saying they are methodically equivalent (although there are similarities, e.g., the Dirichlet Process is a distribut

Sorry if I didn't make that clear: I wasn't saying they are methodically equivalent (although there are similarities, e.g., the Dirichlet Process is a distribution over multinomials, which would be one way to formulate a clustering problem) but that the results can be equivalent, as the cited paper shows.