Word2Vec with Spark MLlib on the 20 Newsgroups dataset