NER-LSTM-CRF
NER-LSTM-CRF copied to clipboard
liu-nlper
大佬请问下基于字的准确率高还是基于词的准确率高啊
中文还是基于字吧，可以加入分词特征
On Thu, Oct 25, 2018 at 2:42 PM yuye2133 [email protected] wrote:
RT
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/liu-nlper/NER-LSTM-CRF/issues/28, or mute the thread https://github.com/notifications/unsubscribe-auth/ASWWiqrLyhIMSTl3jk0s3HtqOEkXvkbHks5uoVzcgaJpZM4X5l5Y .
可以将各个特征向量拼接起来输入到lstm
On Fri, Oct 26, 2018 at 2:37 PM yuye2133 [email protected] wrote:
感谢回复，再问个问题哈，我看到除了用字特征，还有词性特征等等，那是各个特征，分别经过双向lstm，最后再concat起来输入到crf里去的吗？
— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/liu-nlper/NER-LSTM-CRF/issues/28#issuecomment-433302750, or mute the thread https://github.com/notifications/unsubscribe-auth/ASWWiqVhX05xztPEUBbpypi3ctAiurSEks5uoq2wgaJpZM4X5l5Y .