Titlebook: Machine Translation; 17th China Conferenc Jinsong Su,Rico Sennrich Conference proceedings 2021 Springer Nature Singapore Pte Ltd. 2021 arti

懸掛

BUST

Communications in Computer and Information Sciencehttp://image.papertrans.cn/m/image/620772.jpg

無效

Nmda-Receptor

,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.

Femish

前兆

dissent

painkillers

,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.

主動

,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.

Adulterate