... ones from hand-crafted features, and that a Bi-LSTM+CRF-model produces state-of-the art accuracy on scansion of poetry in two languages.
確定! 回上一頁