典型文献
Word-Based Method for Chinese Part-of-Speech via Parallel and Adversarial Network
文献摘要：
Chinese part-of-speech(POS)tagging is an essential task for Chinese downstream natural lan-guage processing tasks.The accuracy of the Chinese POS task will drop dramatically by word-based methods be-cause of the segmentation errors and the word sparsity.Also,there are several Chinese POS tagging sets with dif-ferent criteria.Some of them only have a small-scale an-notated corpus and are hard to train.To this end,we propose a modified word-based transformer neural net-work architecture.Meanwhile,we utilize an adversarial transfer learning method that splits the architecture into shared and private parts.This work directly improves the ability of the word-based model,instead of adopting a joint character-based method.Extensive experiments show that our method achieves state-of-the-art perform-ance on all datasets,and more importantly,our method improves performance effectively for the word-based Chinese sequence labeling task.
文献关键词：
中图分类号：
作者姓名：
HUANG Kaiyu;CAO Jingxiang;LIU Zhuang;HUANG Degen
作者机构：
School of Computer Science and Technology,Dalian University of Technology,Dalian 116024,China;School of Foreign Languages,Dalian University of Technology,Dalian 116024,China;School of Applied Finance,Dongbei University of Finance and Economics,Dalian 116025,China
文献出处：
引用格式：
[1]HUANG Kaiyu;CAO Jingxiang;LIU Zhuang;HUANG Degen-.Word-Based Method for Chinese Part-of-Speech via Parallel and Adversarial Network)[J].电子学报（英文）,2022(02):337-344
A类：
notated
B类：
Word,Based,Method,Chinese,Part,Speech,via,Parallel,Adversarial,Network,speech,POS,tagging,essential,downstream,natural,lan,guage,processing,tasks,accuracy,will,drop,dramatically,by,word,methods,cause,segmentation,errors,sparsity,Also,there,several,ferent,criteria,Some,them,only,have,small,scale,corpus,hard,train,To,this,end,we,propose,modified,transformer,neural,net,architecture,Meanwhile,utilize,adversarial,transfer,learning,that,splits,into,shared,private,parts,This,directly,improves,ability,model,instead,adopting,joint,character,Extensive,experiments,show,our,achieves,state,datasets,more,importantly,performance,effectively,sequence,labeling
AB值：
0.622307
相似文献
机标中图分类号，由域田数据科技根据网络公开资料自动分析生成，仅供学习研究参考。
