作者: BIBLE 時間: 2025-3-21 20:17 作者: browbeat 時間: 2025-3-22 00:59
Semantic-Aware Deep Neural Attention Network for Machine Translation Detection,er for monolingual detection and further explores the semantic consistency relationship for bilingual detection. The experimental results on the Chinese-English machine translation detection task show that our models achieve 83.12% . in the monolingual detection and 85.53% . in the bilingual detecti作者: 欺騙手段 時間: 2025-3-22 07:02
Routing Based Context Selection for Document-Level Neural Machine Translation,l structure information more effectively. At the same time, this structured information selection mechanism will also alleviate the possible problems caused by long-distance encoding. Experimental results show that our method is 2.91 BLEU higher than the Transformer model on the public dataset of ZH作者: Resection 時間: 2025-3-22 12:02 作者: Asseverate 時間: 2025-3-22 15:57 作者: Countermand 時間: 2025-3-22 20:06
Nier Wu,Hongxu Hou,Xiaoning Jia,Xin Chang,Haoran Li作者: morale 時間: 2025-3-22 21:29
Nier Wu,Hongxu Hou,Haoran Li,Xin Chang,Xiaoning Jia作者: cunning 時間: 2025-3-23 04:40 作者: 潛移默化 時間: 2025-3-23 08:23
Yiqi Tong,Yidong Chen,Guocheng Zhang,Jiangbin Zheng,Hongkang Zhu,Xiaodong Shi作者: 生命層 時間: 2025-3-23 13:39 作者: –FER 時間: 2025-3-23 15:05 作者: 忙碌 時間: 2025-3-23 21:10
tuted for the unknown true values. It is natural, then, to ask how the variance–covariance parameters should be estimated. Answering this question is the topic of this chapter. We begin with an answer that applies when the model is a components-of-variance model, for which a method known as quadrati作者: 斷言 時間: 2025-3-24 01:10 作者: Robust 時間: 2025-3-24 03:45 作者: Sleep-Paralysis 時間: 2025-3-24 08:08 作者: intellect 時間: 2025-3-24 12:46 作者: 協(xié)議 時間: 2025-3-24 15:41
,SAU’S Submission for CCMT 2021 Quality Estimation Task,N-ZH). In this task. We follow TransQuest framework which is based on cross-lingual transformers (XLM-R). In order to make the model pay more attention to key words, we use the attention mechanism and gate module to fuse the last hidden state and pooler output of XLM-R model to generate more accurat作者: DOLT 時間: 2025-3-24 20:14 作者: Anthropoid 時間: 2025-3-25 00:21
Low-Resource Neural Machine Translation Based on Improved Reptile Meta-learning Method, global optimal parameters obtained through transfer learning can not effectively adapt to new tasks, which means the problem of local optimum will be caused when training the new task model. Although this problem can be alleviated by optimization-based meta-learning methods, but meta-parameters are作者: 喚起 時間: 2025-3-25 04:04
Semantic Perception-Oriented Low-Resource Neural Machine Translation,aining methods (BERT) uses attention mechanism based on Levenshtein distance (LD) to extract language features, which ignored syntax-related information. In this paper, we proposed a machine translation pre-training method with semantic perception which depend on the traditional position-based model作者: 喊叫 時間: 2025-3-25 10:57
Semantic-Aware Deep Neural Attention Network for Machine Translation Detection,of data collected comes from machine-translated texts rather than native speakers or professional translators, severely reducing the benefit of data scale. Traditional machine translation detection methods generally require human-crafted feature engineering and are difficult to distinguish the fine-作者: AMEND 時間: 2025-3-25 11:57
Routing Based Context Selection for Document-Level Neural Machine Translation,encoding. Usually, the sentence-level representation is incorporated (via attention or gate mechanism) in these methods, which makes them straightforward but rough, and it is difficult to distinguish useful contextual information from noises. Furthermore, the longer the encoding length is, the more 作者: Musculoskeletal 時間: 2025-3-25 19:17
Generating Diverse Back-Translations via Constraint Random Decoding,erformance of Neural Machine Translation (NMT), especially in low-resource scenarios. Previous researches show that diversity of the synthetic source sentences is essential for back-translation. However, the frequently used random methods such as sampling or noised beam search, although can output d作者: Additive 時間: 2025-3-25 22:57
,ISTIC’s Neural Machine Translation System for CCMT’ 2021,chnical Information of China (ISTIC) for the 17th China Conference on Machine Translation (CCMT’ 2021). ISTIC participated in the following four machine translation (MT) evaluation tasks: MT task of Mongolian-to-Chinese daily expressions, MT task of Tibetan-to-Chinese government documents, MT task o作者: Anticoagulant 時間: 2025-3-26 00:27 作者: 和平主義者 時間: 2025-3-26 06:14 作者: maladorit 時間: 2025-3-26 10:58 作者: Intrepid 時間: 2025-3-26 16:01
1865-0929 ober 2021.?.The 10 papers presented in this volume were carefully reviewed and selected from 25 submissions and focus on all aspects of machine translation, including preprocessing, neural machine translation models, hybrid model, evaluation method, and post-editing..978-981-16-7511-9978-981-16-7512-6Series ISSN 1865-0929 Series E-ISSN 1865-0937 作者: INCUR 時間: 2025-3-26 20:37
978-981-16-7511-9Springer Nature Singapore Pte Ltd. 2021作者: 懸掛 時間: 2025-3-26 21:17 作者: BUST 時間: 2025-3-27 03:39
Communications in Computer and Information Sciencehttp://image.papertrans.cn/m/image/620772.jpg作者: 無效 時間: 2025-3-27 06:59 作者: Nmda-Receptor 時間: 2025-3-27 10:09
,SAU’S Submission for CCMT 2021 Quality Estimation Task,n to key words, we use the attention mechanism and gate module to fuse the last hidden state and pooler output of XLM-R model to generate more accurate prediction. In addition, we use the Predictor-Estimator architecture model to integrate with our model to improve the results. Experiments show that this is a simple and effective ensemble method.作者: Femish 時間: 2025-3-27 15:47 作者: 前兆 時間: 2025-3-27 19:29 作者: dissent 時間: 2025-3-28 00:44 作者: painkillers 時間: 2025-3-28 06:07
,BJTU-Toshiba’s Submission to CCMT 2021 QE and APE Task,nthetic data by different data augmentation methods, i.e. forward translation, round-trip translation and multi-source denoising autoencoder. Multi-model ensemble is adopted in both tasks. Experiment results on the development set show high accuracy on both QE and APE tasks, demonstrating the effectiveness of our proposed methods.作者: 主動 時間: 2025-3-28 06:28
,ISTIC’s Neural Machine Translation System for CCMT’ 2021,ective strategies are adopted to improve the quality of translation, such as corpus filtering, back translation, data augmentation, context-based system combination, model averaging, model ensemble, and reranking. The paper presents the system performance under different parameter settings.作者: Adulterate 時間: 2025-3-28 12:51 作者: 按等級 時間: 2025-3-28 17:58
Tumors of the Tongue and Floor of the Mouth,Der Zweck der nachfolgenden Fragen ist es, festzustellen, ob Ihrem Projekt eine brauchbare Zielsetzung zugrunde liegt.作者: conference 時間: 2025-3-28 19:06 作者: 消瘦 時間: 2025-3-29 01:45
Distributive Justice and Desirable Ends of Economic Activity is the welfare of the consumers, public and private. In no sense is mere production as such a proper measure, rather, it has to be production for the ends that people want. Output, income, and consumption are important aims and preconditions for achieving other goals of individuals; that is, they are only a part of what people live for.
