Recurrent models and lower bounds for projective syntactic decoding

Natalie Schluter

doi:10.18653/v1/N19-1022

Recurrent models and lower bounds for projective syntactic decoding

Publikation: Konference artikel i Proceeding eller bog/rapport kapitel › Konferencebidrag i proceedings › Forskning › peer review

Abstract

The current state-of-the-art in neural graphbased parsing uses only approximate decoding at the training phase. In this paper aim to understand this result better. We show how recurrent models can carry out projective maximum spanning tree decoding. This result holds for both current state-of-the-art models for shiftreduce
and graph-based parsers, projective or not. We also provide the first proof on the
lower bounds of projective maximum spanning tree, DAG, and digraph decoding.

Originalsprog	Engelsk
Titel	Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies
Antal sider	10
Vol/bind	Volume 1 (Long and Short Papers)
Forlag	Association for Computational Linguistics
Publikationsdato	2019
Sider	251-260
DOI	https://doi.org/10.18653/v1/N19-1022
Status	Udgivet - 2019

Emneord

Neural Graph-based Parsing
Approximate Decoding
Recurrent Models
Projective Maximum Spanning Tree
Shift-Reduce Parsers

Adgang til dokumentet

10.18653/v1/N19-1022

watermarked_NAACL2019_NatalieSchluterAccepteret manuskript, 258 KB

https://www.aclweb.org/anthology/N19-1022.pdf

Citationsformater

@inproceedings{b4ca3ad8e91f444d990b57eb0b7b7510,

title = "Recurrent models and lower bounds for projective syntactic decoding",

abstract = "The current state-of-the-art in neural graphbased parsing uses only approximate decoding at the training phase. In this paper aim to understand this result better. We show how recurrent models can carry out projective maximum spanning tree decoding. This result holds for both current state-of-the-art models for shiftreduceand graph-based parsers, projective or not. We also provide the first proof on thelower bounds of projective maximum spanning tree, DAG, and digraph decoding.",

keywords = "Neural Graph-based Parsing, Approximate Decoding, Recurrent Models, Projective Maximum Spanning Tree, Shift-Reduce Parsers, Neural Graph-based Parsing, Approximate Decoding, Recurrent Models, Projective Maximum Spanning Tree, Shift-Reduce Parsers",

author = "Natalie Schluter",

year = "2019",

doi = "10.18653/v1/N19-1022",

language = "English",

volume = "Volume 1 (Long and Short Papers)",

pages = "251--260",

booktitle = "Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies",

publisher = "Association for Computational Linguistics",

address = "United States",

}

Recurrent models and lower bounds for projective syntactic decoding. / Schluter, Natalie.
Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies. Bind Volume 1 (Long and Short Papers) Association for Computational Linguistics, 2019. s. 251-260.

Publikation: Konference artikel i Proceeding eller bog/rapport kapitel › Konferencebidrag i proceedings › Forskning › peer review

TY - GEN

T1 - Recurrent models and lower bounds for projective syntactic decoding

AU - Schluter, Natalie

PY - 2019

Y1 - 2019

N2 - The current state-of-the-art in neural graphbased parsing uses only approximate decoding at the training phase. In this paper aim to understand this result better. We show how recurrent models can carry out projective maximum spanning tree decoding. This result holds for both current state-of-the-art models for shiftreduceand graph-based parsers, projective or not. We also provide the first proof on thelower bounds of projective maximum spanning tree, DAG, and digraph decoding.

AB - The current state-of-the-art in neural graphbased parsing uses only approximate decoding at the training phase. In this paper aim to understand this result better. We show how recurrent models can carry out projective maximum spanning tree decoding. This result holds for both current state-of-the-art models for shiftreduceand graph-based parsers, projective or not. We also provide the first proof on thelower bounds of projective maximum spanning tree, DAG, and digraph decoding.

KW - Neural Graph-based Parsing

KW - Approximate Decoding

KW - Recurrent Models

KW - Projective Maximum Spanning Tree

KW - Shift-Reduce Parsers

KW - Neural Graph-based Parsing

KW - Approximate Decoding

KW - Recurrent Models

KW - Projective Maximum Spanning Tree

KW - Shift-Reduce Parsers

UR - https://www.aclweb.org/anthology/N19-1022/

U2 - 10.18653/v1/N19-1022

DO - 10.18653/v1/N19-1022

M3 - Article in proceedings

VL - Volume 1 (Long and Short Papers)

SP - 251

EP - 260

BT - Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies

PB - Association for Computational Linguistics

ER -

Recurrent models and lower bounds for projective syntactic decoding

Abstract

Emneord

Adgang til dokumentet

Fingeraftryk

Citationsformater