CASIA OpenIR  > 模式识别国家重点实验室  > 自然语言处理
A Dynamic Window Neural Network for CCG Supertagging
Wu HJ(吴惠甲); Zhang JJ(张家俊); Zong CQ(宗成庆); Zong CQ(宗成庆)
2017-02
Conference NameAAAI-17
Conference Date2017
Conference Place美国
AbstractCombinatory Category Grammar (CCG) supertagging is a task to assign lexical categories to each word in a sentence. Almost all previous methods use fixed context window sizes as input features. However, it is obvious that different tags usually rely on different context window sizes. These motivate us to build a supertagger with a dynamic window approach, which can be treated as an attention mechanism on the local contexts. Applying dropout on the dynamic filters can be seen as drop on words directly, which is superior to the regular dropout on word embeddings. We use this approach to demonstrate the state-of-the-art CCG supertagging performance on the standard test set.
KeywordSupertagging Dynamic Window Attention Mechanism
Document Type会议论文
Identifierhttp://ir.ia.ac.cn/handle/173211/14506
Collection模式识别国家重点实验室_自然语言处理
Corresponding AuthorZong CQ(宗成庆)
Affiliation中国科学院自动化研究所
Recommended Citation
GB/T 7714
Wu HJ,Zhang JJ,Zong CQ,et al. A Dynamic Window Neural Network for CCG Supertagging[C],2017.
Files in This Item: Download All
File Name/Size DocType Version Access License
A Dynamic Window Neu(925KB)会议论文 开放获取CC BY-NC-SAView Download
Related Services
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[Wu HJ(吴惠甲)]'s Articles
[Zhang JJ(张家俊)]'s Articles
[Zong CQ(宗成庆)]'s Articles
Baidu academic
Similar articles in Baidu academic
[Wu HJ(吴惠甲)]'s Articles
[Zhang JJ(张家俊)]'s Articles
[Zong CQ(宗成庆)]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[Wu HJ(吴惠甲)]'s Articles
[Zhang JJ(张家俊)]'s Articles
[Zong CQ(宗成庆)]'s Articles
Terms of Use
No data!
Social Bookmark/Share
File name: A Dynamic Window Neural Network for CCG Supertagging.pdf
Format: Adobe PDF
All comments (0)
No comment.
 

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.