Distributed Representations of Words and Phrases and their Compositionality
The recently introduced continuous Skip-gram model is an efficient method for learning high-quality distributed vector representations that capture a large number of precise syn...