Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

Stack-Pointer Networks for Dependency Parsing

About

We introduce a novel architecture for dependency parsing: \emph{stack-pointer networks} (\textbf{\textsc{StackPtr}}). Combining pointer networks~\citep{vinyals2015pointer} with an internal stack, the proposed model first reads and encodes the whole sentence, then builds the dependency tree top-down (from root-to-leaf) in a depth-first fashion. The stack tracks the status of the depth-first search and the pointer networks select one child for the word at the top of the stack at each step. The \textsc{StackPtr} parser benefits from the information of the whole sentence and all previously derived subtree structures, and removes the left-to-right restriction in classical transition-based parsers. Yet, the number of steps for building any (including non-projective) parse tree is linear in the length of the sentence just as other transition-based parsers, yielding an efficient decoding algorithm with $O(n^2)$ time complexity. We evaluate our model on 29 treebanks spanning 20 languages and different dependency annotation schemas, and achieve state-of-the-art performance on 21 of them.

Xuezhe Ma, Zecong Hu, Jingzhou Liu, Nanyun Peng, Graham Neubig, Eduard Hovy• 2018

Related benchmarks

TaskDatasetResultRank
Dependency ParsingChinese Treebank (CTB) (test)
UAS92.58
99
Dependency ParsingPenn Treebank (PTB) (test)
LAS95.34
80
Dependency ParsingEnglish PTB Stanford Dependencies (test)
UAS95.87
76
Dependency ParsingUD 2.2 (test)
bg93.04
31
Dependency ParsingCoNLL German 2009 (test)
UAS94.77
25
Dependency ParsingPTB
UAS95.87
24
Dependency ParsingDep. Parse (test)
UAS95.9
23
Dependency ParsingCoNLL Spanish 2009 (test)
UAS90.87
14
Dependency ParsingCoNLL English 2009 (test)
UAS93.25
13
Dependency ParsingCoNLL Czech 2009 (test)
UAS92.83
12
Showing 10 of 23 rows

Other info

Code

Follow for update