method: Encoder_Decoder_v12019-10-30

Authors: Wuheng Xu, Changxu Cheng, Bohan Li´╝îBinghui Xu

Description: We used a Encoder-Decoder structure with BLSTM and attention-based network.We also used some data augments.Finally we trained it on the Chinese datasets all we can find and finetuned it in ReCTs.