We present an efficient approach for clustering a massive stream of textual documents, with particular emphasis on parallelization by the means of multithreaded processing. The underlying cluste