Word2Vec

Word2Vec is a Two Layer Neural Network based Continuous bag of word (CBOW) and Skip-gram architecture that captures the semantic information. It generates the word embedding (mapping of words in a vector space) for a given text corpus. It converts the words into vectors and vectors performs an operation like add, subtract, calculating distance, etc. which preserves the relationship among the words. How are the relationships among words are formed? Word2Vec assigns similar vector representation to the similar words.