Weston

Google researchers led by Tomas Mikolov published word2vec in 2013, a method that efficiently learns dense word embeddings from large text corpora. By training a shallow neural network to predict cont…