Machine Translation Based on Domain Adaptive Language Model

被引：0

作者：

Li, Lingling ^{[1
]}

Chen, Xianlong ^{[1
]}

Xu, Yiling ^{[1
]}

机构：

[1] Dongguan Univ Technol, City Coll, Dongguan, Peoples R China

来源：

2020 16TH INTERNATIONAL CONFERENCE ON COMPUTATIONAL INTELLIGENCE AND SECURITY (CIS 2020) | 2020年

关键词：

machine translation; language model; domain adaptation; adjustable discount; domain interpolation;

D O I：

10.1109/CIS52066.2020.00033

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

This study presents a domain adaptive language model based on adjustable parameters and domain interpolation. Meanwhile, a method of automatically determining test data domain by language model is proposed. Results show that the perplexity of the proposed language model is significantly lower than that of the baseline of KN smoothing language model on the cross-domain test set. In Chinese-English translation, the BLEU value of machine translation evaluation is also significantly higher than that of baseline model.

引用

页码：116 / 120

页数：5

共 9 条

[1] An empirical study of smoothing techniques for language modeling
Chen, SF
Goodman, J
[J]. COMPUTER SPEECH AND LANGUAGE, 1999, 13 (04) : 359 - 394
[2] Chiang David, 2005, P ACL
[3] Guo J., 2014, P 11 BIENN C ASS MAC P 11 BIENN C ASS MAC
[4] Heafield Kenneth, 2013, P 51 ANN M ASS COMP, V2
[5] Koehn P., 2009, Statistical machine translation, DOI DOI 10.1017/CBO9780511815829
[6] Manning Chris, 1999, Foundations of Statistical Natural Language Processing
[7] BLEU: a method for automatic evaluation of machine translation
Papineni, K
Roukos, S
Ward, T
Zhu, WJ
[J]. 40TH ANNUAL MEETING OF THE ASSOCIATION FOR COMPUTATIONAL LINGUISTICS, PROCEEDINGS OF THE CONFERENCE, 2002, : 311 - 318
[8] Stolcke A., 2002, INTERSPEECH, V2
[9] Zens R., 2002, ANN C ART INT ANN C ART INT

← 1 →