Filters








2,605 Hits in 6.2 sec

Neural Quality Estimation with Multiple Hypotheses for Grammatical Error Correction [article]

Zhenghao Liu, Xiaoyuan Yi, Maosong Sun, Liner Yang, Tat-Seng Chua
2021 arXiv   pre-print
However, existing models neglect the possible GEC evidence from different hypotheses. This paper presents the Neural Verification Network (VERNet) for GEC quality estimation with multiple hypotheses.  ...  Grammatical Error Correction (GEC) aims to correct writing errors and help language learners improve their writing skills.  ...  Acknowledgments We thank the reviewers and Shuo Wang for their valuable comments and advice.  ... 
arXiv:2105.04443v1 fatcat:pkthay3t5ja73iij6blynwy5c4

Neural Quality Estimation with Multiple Hypotheses for Grammatical Error Correction

Zhenghao Liu, Xiaoyuan Yi, Maosong Sun, Liner Yang, Tat-Seng Chua
2021 Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies   unpublished
However, existing models neglect the possible GEC evidence from different hypotheses. This paper presents the Neural Verification Network (VERNet) for GEC quality estimation with multiple hypotheses.  ...  Grammatical Error Correction (GEC) aims to correct writing errors and help language learners improve their writing skills.  ...  Acknowledgments We thank the reviewers and Shuo Wang for their valuable comments and advice.  ... 
doi:10.18653/v1/2021.naacl-main.429 fatcat:bm7flpwowjej3jnnshtlokotei

Neural Quality Estimation of Grammatical Error Correction

Shamil Chollampatt, Hwee Tou Ng
2018 Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing  
Our neural quality estimation models for GEC show significant improvements over a strong feature-based baseline.  ...  Grammatical error correction (GEC) systems deployed in language learning environments are expected to accurately correct errors in learners' writing.  ...  Introduction The task of automatically correcting various kinds of errors in written text, termed as grammatical error correction (GEC), is primarily aimed at assisting language learning and providing  ... 
doi:10.18653/v1/d18-1274 dblp:conf/emnlp/ChollampattN18 fatcat:3qblkfwdjnd7lnjjtjumpvsir4

TMU Transformer System Using BERT for Re-ranking at BEA 2019 Grammatical Error Correction on Restricted Track

Masahiro Kaneko, Kengo Hotate, Satoru Katsumata, Mamoru Komachi
2019 Proceedings of the Fourteenth Workshop on Innovative Use of NLP for Building Educational Applications  
Therefore, we propose to finetune BERT on learner corpora with grammatical errors for re-ranking.  ...  grammatical errors.  ...  Acknowledgments We thank Yangyang Xi of Lang-8, Inc. for kindly allowing us to use the Lang-8 learner corpus.  ... 
doi:10.18653/v1/w19-4422 dblp:conf/bea/KanekoHKK19 fatcat:tzzklk2jgjanvcxhm6uhsfpjau

Neural and FST-based approaches to grammatical error correction

Zheng Yuan, Felix Stahlberg, Marek Rei, Bill Byrne, Helen Yannakoudakis
2019 Proceedings of the Fourteenth Workshop on Innovative Use of NLP for Building Educational Applications  
In this paper, we describe our submission to the BEA 2019 shared task on grammatical error correction. We present a system pipeline that utilises both error detection and correction models.  ...  The input text is first corrected by two complementary neural machine translation systems: one using convolutional networks and multi-task learning, and another using a neural Transformer-based system.  ...  Felix Stahlberg and Bill Byrne acknowledge support for this work by the U.K.  ... 
doi:10.18653/v1/w19-4424 dblp:conf/bea/YuanSRBY19 fatcat:a7zt3un3fnge7jugluvguq5e5i

A Neural Grammatical Error Correction System Built On Better Pre-training and Sequential Transfer Learning

Yo Joong Choe, Jiyeon Ham, Kyubyong Park, Yeoil Yoon
2019 Proceedings of the Fourteenth Workshop on Innovative Use of NLP for Building Educational Applications  
Grammatical error correction can be viewed as a low-resource sequence-to-sequence task, because publicly available parallel corpora are limited.  ...  Combined with a context-aware neural spellchecker, our system achieves competitive results in both restricted and low resource tracks in ACL 2019 BEA Shared Task.  ...  Grammatical error correction with neural reinforcement learning.  ... 
doi:10.18653/v1/w19-4423 dblp:conf/bea/ChoeHPY19 fatcat:36o5uczizrh63kfcdggcz5mhxm

A Comprehensive Survey of Grammar Error Correction [article]

Yu Wang, Yuelin Wang, Jie Liu, Zhuo Liu
2020 arXiv   pre-print
Grammar error correction (GEC) is an important application aspect of natural language processing techniques.  ...  Similarly, some performance boosting techniques are adapted from machine translation and are successfully combined with GEC systems for enhancement on the final performance.  ...  • Neural quality estimation Chollampatt and Ng proposed the first neural quality estimation model for GEC and used the quality estimation score as a feature in reranking, which yields statistically  ... 
arXiv:2005.06600v1 fatcat:p4op2mwbefdqtfsewnhrvhcl6q

How to Ask Better Questions? A Large-Scale Multi-Domain Dataset for Rewriting Ill-Formed Questions [article]

Zewei Chu, Mingda Chen, Jing Chen, Miaosen Wang, Kevin Gimpel, Manaal Faruqui, Xiance Si
2019 arXiv   pre-print
We provide human annotations for a subset of the dataset as a quality estimate.  ...  We present a large-scale dataset for the task of rewriting an ill-formed natural language question to a well-formed one.  ...  Acknowledgments We thank Shankar Kumar, Zi Yang, Yiran Zhang, Rahul Gupta, Dekang Lin, Yuchen Lin, Guan-lin Chao, Llion Jones, and Amarnag Subramanya for their helpful discussions and suggestions.  ... 
arXiv:1911.09247v1 fatcat:6dqfmrdrabfptg22qnduibsllq

How to Ask Better Questions? A Large-Scale Multi-Domain Dataset for Rewriting Ill-Formed Questions

Zewei Chu, Mingda Chen, Jing Chen, Miaosen Wang, Kevin Gimpel, Manaal Faruqui, Xiance Si
2020 PROCEEDINGS OF THE THIRTIETH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE AND THE TWENTY-EIGHTH INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE  
We provide human annotations for a subset of the dataset as a quality estimate.  ...  We present a large-scale dataset for the task of rewriting an ill-formed natural language question to a well-formed one.  ...  Acknowledgments We thank Shankar Kumar, Zi Yang, Yiran Zhang, Rahul Gupta, Dekang Lin, Yuchen Lin, Guan-lin Chao, Llion Jones, and Amarnag Subramanya for their helpful discussions and suggestions.  ... 
doi:10.1609/aaai.v34i05.6258 fatcat:mionie2jbvex5a43b76dgr5xri

A Self-Refinement Strategy for Noise Reduction in Grammatical Error Correction [article]

Masato Mita, Shun Kiyono, Masahiro Kaneko, Jun Suzuki, Kentaro Inui
2020 arXiv   pre-print
Existing approaches for grammatical error correction (GEC) largely rely on supervised learning with manually created GEC datasets.  ...  However, there has been little focus on verifying and ensuring the quality of the datasets, and on how lower-quality data might affect GEC performance.  ...  Acknowledgments We thank the Tohoku NLP laboratory members who provided us with their valuable advice.  ... 
arXiv:2010.03155v1 fatcat:6gcv4g5nwnbhfefxfd6s5kg6j4

A Semantic Role-based Approach to Open-Domain Automatic Question Generation

Michael Flor, Brian Riordan
2018 Proceedings of the Thirteenth Workshop on Innovative Use of NLP for Building Educational Applications  
The SRL-based system outperforms the neural system in both average quality and variety of generated questions.  ...  We present an extensive evaluation of the system and compare it to a recent neural network architecture for question generation.  ...  Analysis of NN system errors The patterns of ratings for errorful questions from the neural system differed from the SRL system.  ... 
doi:10.18653/v1/w18-0530 dblp:conf/bea/FlorR18 fatcat:himhlffbsfal7aa46mjgm2ur4y

Diversity-Driven Combination for Grammatical Error Correction [article]

Wenjuan Han, Hwee Tou Ng
2021 arXiv   pre-print
Grammatical error correction (GEC) is the task of detecting and correcting errors in a written text. The idea of combining multiple system outputs has been successfully used in GEC.  ...  To achieve successful system combination, multiple component systems need to produce corrected sentences that are both diverse and of comparable quality.  ...  We thank Weiqi Wang and Yang Song for training the backbone component system, and Muhammad Reza Qorib for helpful feedback and comments.  ... 
arXiv:2110.15149v1 fatcat:3yyzpo2dvrdwbczteqyv6hbtqi

Probabilistic orthographic cues to grammatical category in the brain

Joanne Arciuli, Katie McMahon, Greig de Zubicaray
2012 Brain and Language  
These findings align with an emergentist view of grammatical category processing which results from sensitivity to multiple probabilistic cues. Crown  ...  We report on cues in the way individual English words are spelled, and, for the first time, identify their neural correlates via functional magnetic resonance imaging (fMRI).  ...  Acknowledgments We are grateful to Emily Moseley and Kori Johnson for their assistance with data acquisition.  ... 
doi:10.1016/j.bandl.2012.09.009 pmid:23117157 fatcat:w7g5b5kijngj7ovonaazghe7yi

Disfluencies and Human Speech Transcription Errors

Vicky Zayats, Trang Tran, Richard Wright, Courtney Mansfield, Mari Ostendorf
2019 Interspeech 2019  
A new version of the Switchboard corpus is provided with disfluency annotations for careful speech transcripts, together with results showing the impact of transcription errors on evaluation of automatic  ...  This paper explores contexts associated with errors in transcription of spontaneous speech, shedding light on human perception of disfluencies and other conversational speech phenomena.  ...  We hypothesize that the same will be true for disfluencies in speech that involve repetition or correction.  ... 
doi:10.21437/interspeech.2019-3134 dblp:conf/interspeech/ZayatsTWMO19 fatcat:luqwlwxjqngbbivxysyvbzl2cm

Neural Grammatical Error Correction Systems with Unsupervised Pre-training on Synthetic Data

Roman Grundkiewicz, Marcin Junczys-Dowmunt, Kenneth Heafield
2019 Proceedings of the Fourteenth Workshop on Innovative Use of NLP for Building Educational Applications  
Considerable effort has been made to address the data sparsity problem in neural grammatical error correction.  ...  On the popular CoNLL 2014 test set, we report state-of-theart results of 64.16 M 2 for the submitted system, and 61.30 M 2 for the constrained system trained on the NUCLE and Lang-8 data.  ...  Acknowledgments This work was performed using resources provided by the Cambridge Service for Data Driven Discovery (CSD3) operated by the University of Cambridge Research Computing Service (http: //www.csd3  ... 
doi:10.18653/v1/w19-4427 dblp:conf/bea/GrundkiewiczJH19 fatcat:qkl7zydobzg5xdverinpjhckhu
« Previous Showing results 1 — 15 out of 2,605 results