北京语言大学-语言资源高精尖创新中心在研人员论文成果语言资源高精尖创新中心在研人员发表的相关论文
THIRTY-THIRD AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE / THIRTY-FIRST INNOVATIVE APPLICATIONS OF AR..., pp.451-458, (2019)
The extensive experiments demonstrate that our method significantly outperforms the strong baseline models in translation quality and sharply reduces the under-translation cases of these high-entropy words
Cited by1BibtexViews126
0
0
EMNLP, (2018): 533-542
The past several years have witnessed the rapid development of neural machine translation, which investigates the use of neural networks to model the translation process
Cited by98BibtexViews336DOI
0
0
Yong Cheng,Zhaopeng Tu,Fandong Meng, Junjie Zhai,Yang Liu
meeting of the association for computational linguistics, (2018)
We have proposed adversarial stability training to improve the robustness of Neural machine translation models
Cited by77BibtexViews82DOI
0
0
meeting of the association for computational linguistics, (2018)
We have described the average attention network that considerably alleviates the decoding bottleneck of the neural Transformer
Cited by36BibtexViews63DOI
0
0
EMNLP, pp.2955-2960, (2018)
We have proposed three effective strategies to improve the universal one-to-many multilingual translation, including special label initialization, language-dependent positional embedding and a new parameter-sharing mechanism
Cited by13BibtexViews317
0
0
IJCAI, (2018)
We have proposed a method to incorporate a phrase translation table as recommendation memory into Neural Machine Translation systems to alleviate the problem that the NMT system is opt to generate fluent but unfaithful translations
Cited by12BibtexViews52
0
0
EMNLP, pp.391-400, (2018)
The extensive experiments on Chinese-to-English and English-toGerman translation tasks demonstrate that our method significantly outperforms the strong baseline models in translation quality, especially in handling the troublesome words
Cited by10BibtexViews87
0
0
Le Sun,Ben He,Kai Hui, Cancan Jin
ACL, pp.1088-1097, (2018)
This study aims at addressing the promptindependent automated essay scoring, where no rated essay for the target prompt is available
Cited by10BibtexViews71
0
0
NATURAL LANGUAGE PROCESSING TECHNIQUES FOR EDUCATIONAL APPLICATIONS, pp.42-51, (2018)
This study describes the Natural Language Processing Techniques for Educational Applications 2018 shared task for Chinese grammatical error diagnosis, including task design, data preparation, performance metrics, and evaluation results
Cited by7BibtexViews69
0
0
Hongkai Ren, Liner Yang,Endong Xun
Lecture Notes in Artificial Intelligence, (2018): 401-410
We explored a seq2seq model based entirely on convolutional neural network
Cited by6BibtexViews163DOI
0
0
Chunhua Liu,Shan Jiang, Hainan Yu,Dong Yu
Lecture Notes in Artificial Intelligence, (2018): 131-143
Experimental results on SNLI dataset show that the Multi-turn Inference Matching Network model is on par with the state-of-the-art models
Cited by3BibtexViews205DOI
0
0
Yan Zhao,Lu Liu, Chunhua Liu, Ruoyao Yang,Dong Yu
Lecture Notes in Artificial Intelligence, (2018): 51-63
In this work we propose a framework consisting of a Generator and a Reward Manager to solve the Story Ending Generation problem
Cited by2BibtexViews172DOI
0
0
Longfei Yang, Yanlu Xie,Jinsong Zhang
Interspeech, pp.352-356, (2018)
The Convolutional Neural Network-Bidirectional Long Short-Term Memory with attention mechanism achieved the five-tone error rate of 9.3% which is the best result comparing to other models
Cited by1BibtexViews50
0
0
Longfei Yang, Yanlu Xie,Jinsong Zhang
international conference on signal processing, (2018)
We improved the performance through passing the output of Convolutional Neural Network which could reduce spectral variation, to DBLSTM
Cited by0BibtexViews52DOI
0
0
Qi Zhang, Chong Cao, Tiantian Li,Yanlu Xie,Jinsong Zhang
international conference on signal processing, (2018)
When the pitch information was added to the Fbank and MelFrequency Cepstral Coefficient features, the pitch range prediction performance was not significantly improved if the acoustic input length was more than 300ms
Cited by0BibtexViews52DOI
0
0
2018 11th International Symposium on Chinese Spoken Language Processing (ISCSLP), pp.444-448, (2018)
A phone embedding approach based on Siamese networks was proposed to provide more accurate evaluation for L2 learners’ mispronunciation at phone level
Cited by0BibtexViews165DOI
0
0
2018 11th International Symposium on Chinese Spoken Language Processing (ISCSLP), pp.349-353, (2018)
Based on the hypothesis that the spectral structure of speech is correlated with the overall pitch range, the present study proposed a novel method for estimating pitch range from the spectral structure of a brief speech input, using deep learning techniques
Cited by0BibtexViews147DOI
0
0
2018 11th International Symposium on Chinese Spoken Language Processing (ISCSLP), pp.439-443, (2018)
The present study proposed a quantitative method to estimate the prosodic strength of each syllable in Mandarin continuous speech from three acoustic measures
Cited by0BibtexViews157DOI
0
0
Ju Lin,Wei Zhang, Linxuan Wei, Yanlu Xie,Jinsong Zhang
2018 11th International Symposium on Chinese Spoken Language Processing (ISCSLP), pp.170-174, (2018)
Explicit and implicit combinations with λ=0 were the standard student-teacher model, which utilized soft targets generated from Convolutional Neural Network to guide the training of Deep Neural Network
Cited by0BibtexViews151DOI
0
0
2018 11th International Symposium on Chinese Spoken Language Processing (ISCSLP), pp.175-179, (2018)
This paper combined perceptual experiments with phones classification experiments to verify the acoustic landmarks of Mandarin alveolar-palatal consonants
Cited by0BibtexViews156DOI
0
0
小科