Filters








297 Hits in 1.5 sec

Page 243 of Computational Linguistics Vol. 25, Issue 2 [page]

1999 Computational Linguistics  
Table 2 Supertag ambiguity with and without the use of structural constraints.  ...  Table 2 indicates the decrease in supertag ambiguity for 2,012 WSJ sentences (48,763 words) by using the structural constraints relative to the supertag ambigu- ity without the structural constraints.”  ... 

Multi-tagging for lexicalized-grammar parsing

James R. Curran, Stephen Clark, David Vadas
2006 Proceedings of the 21st International Conference on Computational Linguistics and the 44th annual meeting of the ACL - ACL '06  
We describe a multi-tagging approach which maintains a suitable level of lexical category ambiguity for accurate and efficient CCG parsing.  ...  Although POS tagging accuracy seems high, maintaining some POS tag ambiguity in the language processing pipeline results in more accurate CCG supertagging.  ...  For lexicalized grammars, such as CCG and TAG, the motivation for using a multi-tagger to assign the elementary structures (supertags) is more compelling.  ... 
doi:10.3115/1220175.1220263 dblp:conf/acl/CurranCV06 fatcat:fxumcithmrguffp4rkxubceafm

Analysis of the Difficulties in Chinese Deep Parsing

Kun Yu, Yusuke Miyao, Takuya Matsuzaki, Xiangli Wang, Jun'ichi Tsujii
2011 International Workshop/Conference on Parsing Technologies  
This paper discusses the difficulties in Chinese deep parsing, by comparing the accuracy of a Chinese HPSG parser to the accuracy of an English HPSG parser and the commonly used Chinese syntactic parsers  ...  Moreover, the inherent ambiguities caused by verbal coordination and relative clauses make semantic analysis of Chinese more difficult than the syntactic analysis of Chinese.  ...  The feature forest model provides a solution to the problem of probabilistic modeling of complex data structures.  ... 
dblp:conf/iwpt/YuMMWT11 fatcat:ljbapnrtgfejxaxevgizxvdhza

Transition-Based Dependency Parsing Exploiting Supertags

Hiroki Ouchi, Kevin Duh, Hiroyuki Shindo, Yuji Matsumoto
2016 IEEE/ACM Transactions on Audio Speech and Language Processing  
In this work, we present a supertag design framework that allows us to instantiate various supertag sets based on the dependency structures.  ...  Using this framework, we instantiate various supertag sets and utilize them as features in transition-based dependency parsing systems.  ...  Unigram, Bigram, and Structural features are based on the features used in [18] and [19] with some modifications, which we call Base Features.  ... 
doi:10.1109/taslp.2016.2598310 fatcat:u2moqc2njzborah5ifa2idcgse

A Simple Approach for HPSG Supertagging Using Dependency Information

Yao-zhong Zhang, Takuya Matsuzaki, Jun'ichi Tsujii
2010 North American Chapter of the Association for Computational Linguistics  
In a supertagging task, sequence labeling models are commonly used. But their limited ability to model long-distance information presents a bottleneck to make further improvements.  ...  In this paper, we modeled this long-distance information in dependency formalism and integrated it into the process of HPSG supertagging.  ...  The first author was supported by The University of Tokyo Fellowship (UT-Fellowship). 4 Tested on section 23 (2291 sentences) using an AMD Opteron 2.4GHz server, given all supertag candidates.  ... 
dblp:conf/naacl/ZhangMT10 fatcat:6zu7iyzgajdnbav2plr23jcjae

TAG Parsing with Neural Networks and Vector Representations of Supertags

Jungo Kasai, Bob Frank, Tom McCoy, Owen Rambow, Alexis Nasr
2017 Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing  
The shift-reduce parsing model eschews lexical information entirely, and uses only the 1-best supertags to parse a sentence, providing further support for the claim that supertagging is "almost parsing  ...  We present supertagging-based models for Tree Adjoining Grammar parsing that use neural network architectures and dense vector representation of supertags (elementary trees) to achieve state-of-the-art  ...  Using these features as input, we build a two-layer feed-forward network to predict the action for the parser to take.  ... 
doi:10.18653/v1/d17-1180 dblp:conf/emnlp/KasaiFMRN17 fatcat:rmxanvnmlzgsrbrz4ycrhm7qji

Deconstructing Supertagging into Multi-Task Sequence Prediction

Zhenqi Zhu, Anoop Sarkar
2019 Proceedings of the 23rd Conference on Computational Natural Language Learning (CoNLL)  
Supertagging is a sequence prediction task where each word is assigned a complex syntactic structure called a supertag.  ...  tasks, which can best represent the structural information of each supertag.  ...  Structural features of supertags are heavily used in pre-neural statistical parsing methods (Bangalore et al., 2009 ) and proved to be useful.  ... 
doi:10.18653/v1/k19-1002 dblp:conf/conll/ZhuS19 fatcat:lpnyd63yp5bgjhnyw4hoe6ia4q

Improved CCG Parsing with Semi-supervised Supertagging

Mike Lewis, Mark Steedman
2014 Transactions of the Association for Computational Linguistics  
We also give a detailed error analysis demonstrating where using embeddings outperforms traditional feature sets, and showing how including POS features can decrease accuracy.  ...  The use of word embeddings enables our model to better generalize from the labelled data, and allows us to accurately assign lexical categories without depending on a POS-tagger.  ...  CCG categories contain much of the hierarchical structure needed for parsing, giving a simpler way to improve a parser using embeddings.  ... 
doi:10.1162/tacl_a_00186 fatcat:l2tki4kgjjctbd3cwivmwcysna

Improving Dependency Parsers with Supertags

Hiroki Ouchi, Kevin Duh, Yuji Matsumoto
2014 Proceedings of the 14th Conference of the European Chapter of the Association for Computational Linguistics, volume 2: Short Papers  
In this paper, we investigate richer features based on supertags, which represent lexical templates extracted from dependency structure annotated corpus.  ...  Then, we propose a transition-based dependency parser that incorporates the predictions from a CRF-based supertagger as new features.  ...  For the unigram features, we use information within a single partial structure, such as conjunction of head word and its supertag (w p s p ), conjunction of head word's POS tag and its supertag (t p s  ... 
doi:10.3115/v1/e14-4030 dblp:conf/eacl/OuchiDM14 fatcat:fldbauaeibfyzdekc3srzpknwy

LSTM CCG Parsing

Mike Lewis, Kenton Lee, Luke Zettlemoyer
2016 Proceedings of the 2016 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies  
We demonstrate that a state-of-the-art parser can be built using only a lexical tagging model and a deterministic grammar, with no explicit model of bi-lexical dependencies.  ...  By running the LSTM on a GPU, we are able to parse over 2600 sentences per second while improving state-of-the-art accuracy by 1.1 F1 in domain and up to 4.5 F1 out of domain.  ...  This research was supported in part by the NSF (IIS-1252835), DARPA under the DEFT program through the AFRL (FA8750-13-2-0019), an Allen Distinguished Investigator Award, and a gift from Google.  ... 
doi:10.18653/v1/n16-1026 dblp:conf/naacl/LewisLZ16 fatcat:cuvcrkkwwzgdfi7t5v47icpige

Supertagging: Introduction, learning, and application [article]

Taraka Rama K
2014 arXiv   pre-print
In the beginning, the scholars used small training datasets and somewhat na\"ive smoothing techniques to learn the probability distributions of supertags.  ...  Supertagging is an approach originally developed by Bangalore and Joshi (1999) to improve the parsing efficiency.  ...  The author also uses supertag based features modeled in the lines of Gildea and Palmer (2002) 4 .  ... 
arXiv:1412.6264v1 fatcat:ffeqidp475fndknqilehy7ol5m

Disambiguation of Super Parts of Speech (or Supertags): Almost Parsing [article]

Aravind K. Joshi, B. Srinivas
1994 arXiv   pre-print
In a lexicalized grammar formalism such as Lexicalized Tree-Adjoining Grammar (LTAG), each lexical item is associated with at least one elementary structure (supertag) that localizes syntactic and semantic  ...  We present techniques for disambiguating supertags using local information such as lexical preference and local lexical dependencies.  ...  Since LTAGs are lexicalized, we are presented with a novel opportunity to eliminate or substantially reduce the supertag assignment ambiguity by using local information such as local lexical dependencies  ... 
arXiv:cmp-lg/9410029v1 fatcat:2rg4wto4n5cp3dcmn2mydly2ia

A* CCG Parsing with a Supertag and Dependency Factored Model

Masashi Yoshikawa, Hiroshi Noji, Yuji Matsumoto
2017 Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)  
We propose a new A* CCG parsing model in which the probability of a tree is decomposed into factors of CCG categories and its syntactic dependencies both defined on bi-directional LSTMs.  ...  Our factored model allows the precomputation of all probabilities and runs very efficiently, while modeling sentence structures explicitly via dependencies.  ...  Zhang et al. (2010) use the syntactic dependencies in a different way, and show that dependency-based features are useful for predicting HPSG supertags.  ... 
doi:10.18653/v1/p17-1026 dblp:conf/acl/YoshikawaNM17 fatcat:utphmslhhvdsxkbhu5sts263mq

Efficient HPSG Parsing with Supertagging and CFG-Filtering

Takuya Matsuzaki, Yusuke Miyao, Jun'ichi Tsujii
2007 International Joint Conference on Artificial Intelligence  
Those lexical entry sequences that passed through the CFG-filter are combined into parse trees by using a simple shift-reduce parsing algorithm, in which structural ambiguities are resolved using a classifier  ...  Recent research has shown that supertagging is a key technology to improve both the speed and accuracy of lexicalized grammar parsing.  ...  The abstraction of the feature structures is specified by means of a restrictor [Shieber, 1985] on the feature structures; a restrictor takes a feature structure and overwrites several feature values  ... 
dblp:conf/ijcai/MatsuzakiMT07 fatcat:allx6tykc5hn7apnf743uvjhha

Structure-guided supertagger learning

YAO-ZHONG ZHANG, TAKUYA MATSUZAKI, JUN'ICHI TSUJII
2012 Natural Language Engineering  
Supertagging is an important technique for deep syntactic analysis. A supertagger is usually trained independently of the parser using a sequence labeling method.  ...  In this paper, we propose a forest-guided supertagger training method to alleviate this problem by incorporating global grammar constraints into the supertagging process using a CFGfilter.  ...  Table 1 : 1 Feature templates used for supertagging models.  ... 
doi:10.1017/s1351324912000034 fatcat:bbxg7zv3vzamfmxxsu3mgo363i
« Previous Showing results 1 — 15 out of 297 results