26,958 Hits in 4.4 sec

Applying co-training methods to statistical parsing

Anoop Sarkar
2001 Second meeting of the North American Chapter of the Association for Computational Linguistics on Language technologies 2001 - NAACL '01   unpublished
Co-Training, previously used for classifiers with 2/3 labels, was extended to the complex problem of statistical parsing.  ...  Attachment: Apply substitution or adjunction to combine the elementary trees to form the parse. Training a Statistical Parser How should the parameters (e.g., rule probabilities) be chosen?  ... 
doi:10.3115/1073336.1073359 fatcat:5de4ss4qvfagfhqttgzdsoqvgm

Learning to Parse on Aligned Corpora (Rough Diamond) [chapter]

Cezary Kaliszyk, Josef Urban, Jiří Vyskočil
2015 Lecture Notes in Computer Science  
In this work we start to address this problem by developing approximate probabilistic parsing techniques that autonomously train disambiguation on large corpora.  ...  methods self-improve based on such semantic feedback.  ...  The resulting grammar trees are again transformed back into a HOL parse tree, to which HOL parsing and typechecking is applied as an additional filter.  ... 
doi:10.1007/978-3-319-22102-1_15 fatcat:smfqlcls2vbivmzanpydsmo6mm

Towards Silver Standard Dependency Treebank of Urdu Tweets

2021 International Journal of Advanced Trends in Computer Science and Engineering  
To overcome this deficiency of hand-annotated corpus, researchers have focused their attention on semi-automatic corpus annotation methods.  ...  This paper describes the experiments carried out using semi-automatic methods like self-training and co-training in an attempt for creating silver-standard dependency treebank of Urdu tweets.  ...  A multi-iterative self-training method was applied to Hindi by [13] for improving training domain parsing accuracy.  ... 
doi:10.30534/ijatcse/2021/1501032021 fatcat:w5wrrbnoxndk5d3vh3bioahn64

Combining discriminative re-ranking and co-training for parsing Mandarin speech transcripts

Wen Wang
2009 2009 IEEE International Conference on Acoustics, Speech and Signal Processing  
Discriminative reranking has been able to significantly improve parsing performance, and co-training has proven to be an effective weakly supervised learning algorithm to bootstrap parsers from a small  ...  We show that combining discriminative reranking and co-training could improve the F-measure by 1.8%-2% absolute compared to cotraining two state-of-the-art Chinese parsers without reranking, for parsing  ...  The authors thank Mary Harper and Zhongqiang Huang for discussions on Chinese parsing and discriminative reranking.  ... 
doi:10.1109/icassp.2009.4960681 dblp:conf/icassp/Wang09 fatcat:jnzwnbjqtvdpxe4oaef7qqlira

Self-Training for Enhancement and Domain Adaptation of Statistical Parsers Trained on Small Datasets

Roi Reichart, Ari Rappoport
2007 Annual Meeting of the Association for Computational Linguistics  
Creating large amounts of annotated data to train statistical PCFG parsers is expensive, and the performance of such parsers declines when training and test data are taken from different domains.  ...  This is the first time that self-training with small labeled datasets is applied successfully to these tasks. We were also able to formulate a characterization of when selftraining is valuable.  ...  We would like to thank Dan Roth for his constructive comments on this paper.  ... 
dblp:conf/acl/ReichartR07a fatcat:jt7g6xe6xrfbtf7xy6pqkkpayq

Semi-Supervised Methods for Out-of-Domain Dependency Parsing [article]

Juntao Yu
2018 arXiv   pre-print
To bridge the performance gap between in-domain and out-of-domain, this thesis investigates three semi-supervised techniques for out-of-domain dependency parsing, namely co-training, self-training and  ...  The comparison between those techniques shows that self-training works equally well as co-training on out-of-domain parsing, while dependency language models can improve both in- and out-of-domain accuracies  ...  Agreement Based Co-training In this work, we apply an agreement based co-training to out-of-domain dependency parsing.  ... 
arXiv:1810.02100v1 fatcat:7yrmkamgqzbdbp36zuxzxfj6ku

Online Co-regularized Algorithms [chapter]

Tom de Ruijter, Evgeni Tsivtsivadze, Tom Heskes
2012 Lecture Notes in Computer Science  
We demonstrate that by sequentially co-regularizing prediction functions on unlabeled data points, our algorithm provides improved performance in comparison to supervised methods on several UCI benchmarks  ...  The presented algorithm is particularly applicable to learning tasks where large amounts of (unlabeled) data are available for training.  ...  It generates all parses allowed by its grammar and applies a set of built-in heuristics to predict goodness of the parses.  ... 
doi:10.1007/978-3-642-33492-4_16 fatcat:flirzx2pejfpdgd65pcs5bsqme

Page 276 of Computational Linguistics Vol. 30, Issue 3 [page]

2004 Computational Linguistics  
Applying co-training methods to statistical parsing. In Proceedings of the Second Meeting of the North American Association for Computational Linguistics, Pittsburgh, pages 175-182, June.  ...  Limitations of co-training for natural language learning from large datasets.  ... 

Japanese dependency parsing using co-occurrence information and a combination of case elements

Takeshi Abekawa, Manabu Okumura
2006 Proceedings of the 21st International Conference on Computational Linguistics and the 44th annual meeting of the ACL - ACL '06  
In this paper, we present a method that improves Japanese dependency parsing by using large-scale statistical information.  ...  The results of an experiment in which our method was used to rerank the results obtained using an existing machine learning based parsing method showed that our method can improve the accuracy of the results  ...  Our method takes into account two types of information, not considered in previous statistical (machine learning based) parsing methods.  ... 
doi:10.3115/1220175.1220280 dblp:conf/acl/AbekawaO06 fatcat:5warpx7owzao5obmphbe3inyga

Domain Adaptation for Dependency Parsing via Self-Training

Juntao Yu, Mohab Elkaref, Bernd Bohnet
2015 Proceedings of the 14th International Conference on Parsing Technologies  
We improve parsing accuracy for out-of-domain texts with a self-training approach that uses confidence-based methods to select additional training samples.  ...  We compare two confidence-based methods: The first method uses the parse score of the employed parser to measure the confidence into a parse tree.  ...  Acknowledgments We would like to thank John Barnden for discussions and comments as well as the anonymous reviewers for their helpful reviews.  ... 
doi:10.18653/v1/w15-2201 dblp:conf/iwpt/YuEB15 fatcat:poy6w2srhrcmdh3peosn2m4oly

Ambiguity-aware Ensemble Training for Semi-supervised Dependency Parsing

Zhenghua Li, Min Zhang, Wenliang Chen
2014 Proceedings of the 52nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)  
Experimental results on benchmark data show that our method significantly outperforms the baseline supervised parser and other entire-tree based semi-supervised methods, such as self-training, co-training  ...  This paper proposes a simple yet effective framework for semi-supervised dependency parsing at entire tree level, referred to as ambiguity-aware ensemble training.  ...  Acknowledgments The authors would like to thank the critical and insightful comments from our anonymous reviewers.  ... 
doi:10.3115/v1/p14-1043 dblp:conf/acl/Li0C14 fatcat:aaufbinrwncnhoxu7yd46we2re

Corpus Variation and Parser Performance

Daniel Gildea
2001 Conference on Empirical Methods in Natural Language Processing  
Most work in statistical parsing has focused on a single corpus: the Wall Street Journal portion of the Penn Treebank.  ...  This leads us to a technique for pruning parameters to reduce the size of the parsing model.  ...  Introduction The past several years have seen great progress in the eld of natural language parsing, through the use of statistical methods trained using large corpora of hand-parsed training data.  ... 
dblp:conf/emnlp/Gildea01 fatcat:rcgxnzkzkrd63hv6v2w6mqgfnq

SynWMD: Syntax-aware Word Mover's Distance for Sentence Similarity Evaluation [article]

Chengwei Wei, Bin Wang, C.-C. Jay Kuo
2022 arXiv   pre-print
First, a weighted graph is built upon the word co-occurrence statistics extracted from the syntactic parse trees of sentences. The importance of each word is inferred from graph connectivities.  ...  An improved WMD method using the syntactic parse tree, called Syntax-aware Word Mover's Distance (SynWMD), is proposed to address these two shortcomings in this work.  ...  Difference between parse tree and linear context in SWF. SWF collects co-occurrence statistics from dependency parse trees, which are well-organized structures of sentences.  ... 
arXiv:2206.10029v1 fatcat:gwjaprmbmffpvpso3bxs32h2re

Neural Semantic Parsing by Character-based Translation: Experiments with Abstract Meaning Representations [article]

Rik van Noord, Johan Bos
2017 arXiv   pre-print
We evaluate the character-level translation method for neural semantic parsing on a large corpus of sentences annotated with Abstract Meaning Representations (AMRs).  ...  57.4; (iv) optimizing the training process by using pre-training and averaging a set of models increases performance to 58.7; (v) adding silver-standard training data obtained by an off-the-shelf parser  ...  Acknowledgements First of all we would like to thank Antonio Toral and Lasha Abzianidze for helpful discussion on neural AMR parsing and machine translation.  ... 
arXiv:1705.09980v2 fatcat:wzy6ud6gxjbu5mldcsh7rvf5wi

A Survey of Syntactic-Semantic Parsing Based on Constituent and Dependency Structures [article]

Meishan Zhang
2020 arXiv   pre-print
Constituent parsing is majorly targeted to syntactic analysis, and dependency parsing can handle both syntactic and semantic analysis.  ...  The parsing community includes many tasks, which are difficult to be covered fully. Here we focus on two of the most popular formalizations of parsing: constituent parsing and dependency parsing.  ...  Kawahara [150] apply co-training to constituent parsing, which is similar to self-training but difference in that the example selection is performed by two parsers.  ... 
arXiv:2006.11056v1 fatcat:pd22rciuxzdc5kvghaapjjyg3u
« Previous Showing results 1 — 15 out of 26,958 results