Using Word2vec's center vector and context vector to analysis the collocation relations between Chinese words, and greedily want to extract some dependency relations in sentence (but not so successful).