Language Parsing
Overview
Language parsing group at Natural Language Processing Lab
are working on Chinese Word Segmentation, Part-of-Speech Tagging,
Constituent-based Syntactic Parsing, and Semantic Role Labeling.
People
Chinese Word Segmentation & Part-of-Speech
Tagging
- Chinese word segmentation is defined to split a raw
sentence into words with appropriate delimiters. Part-of-Speech tagging
is defined to assign appropriate Part-of-Speech tags to words in
sentences. Research at the lab focuses on two directions: 1) how to
apply machine learning methods to get better practical-world systems; 2)
how to combine multiple resources (maybe) with different annotation
standards.
- Software: to be available
Constituent-based Syntactic Parsing
- Constituent-based Syntactic parsing aims to assign
parse tree(s) to input sentences. Generally, input sentences are
preprocessed with word segmentation (if for asian languages, ex.
Chinese) and part-of-speech tagging. Research at the lab focues on
syntactic parsing for Chinese.
- Software: to be available
Semantic Role Labeling
- Semantic role labeling is used to identify
predicate-argument structure(s) in a sentence. Generally semantic role
labeling uses syntactic parse tree(s) as input. The research focus of
our lab is to jointly learn syntactic parsing and semantic role
labeling. Moreover, we are seeking extensive applications of semantic
role labeling.
- Software: to be available
Publication
Muhua Zhu, Jingbo Zhu, and Minghan Hu, Better
Automatic Treebank Conversion Using A Feature-Based Approach.
ACL 2011 (short paper).
Muhua Zhu, Jingbo Zhu, and Tong Xiao
Automatic Treebank Conversion via Informed Decoding - A Case Study
on Chinese Treebanks.
ACM Transactions on Asian Language Information Processing, Speical Issue on
Chinese Language Processing, 2011.
Muhua Zhu, Jingbo Zhu.
Automatic Treebank Conversion via Informed Decoding. Coling 2010 (poster).
Muhua Zhu, Jingbo Zhu, Tong Xiao.
Heterogeneous Parsing via Collaborative Decoding. Coling 2010 (regular).
Haibo Ding, Muhua Zhu, Jingbo Zhu.
A Reranking Method for Syntactic Parsing with Heterogeneous Treebanks.
Proceedings of the 6th IEEE International Conference on
Natural Language Processing and Knowledge Engineering (IEEE NLP-KE'10) .
Muhua Zhu, Huizhen Wang, and Jingbo Zhu, Label
Correspondence Learning for Part-of-Speech Annotation Transformation.
Proceedings of the 18th ACM International Conference on Information and
Knowledge Management ( CIKM'09, short paper).
|