Joint Transition-Based Models for Morpho-Syntactic Parsing: Parsing Strategies for MRLs and a Case Study from Modern Hebrew
In standard NLP pipelines, morphological analysis and disambiguation (MA&D) precedes syntactic and semantic downstream tasks. However, for languages with complex and ambiguous word-internal structure, known as morphologically rich languages (MRLs), it has been hypothesized...
Main Authors: | , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
The MIT Press
2019-11-01
|
Series: | Transactions of the Association for Computational Linguistics |
Online Access: | https://www.mitpressjournals.org/doi/abs/10.1162/tacl_a_00253 |
id |
doaj-1b5e36853e6b4171a4e62028d0858dc6 |
---|---|
record_format |
Article |
spelling |
doaj-1b5e36853e6b4171a4e62028d0858dc62020-11-25T02:43:32ZengThe MIT PressTransactions of the Association for Computational Linguistics2307-387X2019-11-017334810.1162/tacl_a_00253Joint Transition-Based Models for Morpho-Syntactic Parsing: Parsing Strategies for MRLs and a Case Study from Modern HebrewMore, AmirSeker, AmitBasmova, VictoriaTsarfaty, Reut In standard NLP pipelines, morphological analysis and disambiguation (MA&D) precedes syntactic and semantic downstream tasks. However, for languages with complex and ambiguous word-internal structure, known as morphologically rich languages (MRLs), it has been hypothesized that syntactic context may be crucial for accurate MA&D, and vice versa. In this work we empirically confirm this hypothesis for Modern Hebrew, an MRL with complex morphology and severe word-level ambiguity, in a novel transition-based framework. Specifically, we propose a joint morphosyntactic transition-based framework which formally unifies two distinct transition systems, morphological and syntactic, into a single transition-based system with joint training and joint inference. We empirically show that MA&D results obtained in the joint settings outperform MA&D results obtained by the respective standalone components, and that end-to-end parsing results obtained by our joint system present a new state of the art for Hebrew dependency parsing. https://www.mitpressjournals.org/doi/abs/10.1162/tacl_a_00253 |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
More, Amir Seker, Amit Basmova, Victoria Tsarfaty, Reut |
spellingShingle |
More, Amir Seker, Amit Basmova, Victoria Tsarfaty, Reut Joint Transition-Based Models for Morpho-Syntactic Parsing: Parsing Strategies for MRLs and a Case Study from Modern Hebrew Transactions of the Association for Computational Linguistics |
author_facet |
More, Amir Seker, Amit Basmova, Victoria Tsarfaty, Reut |
author_sort |
More, Amir |
title |
Joint Transition-Based Models for Morpho-Syntactic Parsing: Parsing Strategies for MRLs and a Case Study from Modern Hebrew |
title_short |
Joint Transition-Based Models for Morpho-Syntactic Parsing: Parsing Strategies for MRLs and a Case Study from Modern Hebrew |
title_full |
Joint Transition-Based Models for Morpho-Syntactic Parsing: Parsing Strategies for MRLs and a Case Study from Modern Hebrew |
title_fullStr |
Joint Transition-Based Models for Morpho-Syntactic Parsing: Parsing Strategies for MRLs and a Case Study from Modern Hebrew |
title_full_unstemmed |
Joint Transition-Based Models for Morpho-Syntactic Parsing: Parsing Strategies for MRLs and a Case Study from Modern Hebrew |
title_sort |
joint transition-based models for morpho-syntactic parsing: parsing strategies for mrls and a case study from modern hebrew |
publisher |
The MIT Press |
series |
Transactions of the Association for Computational Linguistics |
issn |
2307-387X |
publishDate |
2019-11-01 |
description |
In standard NLP pipelines, morphological analysis and disambiguation (MA&D) precedes syntactic and semantic downstream tasks. However, for languages with complex and ambiguous word-internal structure, known as morphologically rich languages (MRLs), it has been hypothesized that syntactic context may be crucial for accurate MA&D, and vice versa. In this work we empirically confirm this hypothesis for Modern Hebrew, an MRL with complex morphology and severe word-level ambiguity, in a novel transition-based framework. Specifically, we propose a joint morphosyntactic transition-based framework which formally unifies two distinct transition systems, morphological and syntactic, into a single transition-based system with joint training and joint inference. We empirically show that MA&D results obtained in the joint settings outperform MA&D results obtained by the respective standalone components, and that end-to-end parsing results obtained by our
joint system present a new state of the art for Hebrew dependency parsing. |
url |
https://www.mitpressjournals.org/doi/abs/10.1162/tacl_a_00253 |
work_keys_str_mv |
AT moreamir jointtransitionbasedmodelsformorphosyntacticparsingparsingstrategiesformrlsandacasestudyfrommodernhebrew AT sekeramit jointtransitionbasedmodelsformorphosyntacticparsingparsingstrategiesformrlsandacasestudyfrommodernhebrew AT basmovavictoria jointtransitionbasedmodelsformorphosyntacticparsingparsingstrategiesformrlsandacasestudyfrommodernhebrew AT tsarfatyreut jointtransitionbasedmodelsformorphosyntacticparsingparsingstrategiesformrlsandacasestudyfrommodernhebrew |
_version_ |
1724768640438370304 |