7 Hits in 4.7 sec

An Improved Patent Machine Translation System Using Adaptive Enhancement for NTCIR-10 PatentMT Task

Hai Zhao, Jingyi Zhang, Masao Utiyama, Eiichiro Sumita
2013 NTCIR Conference on Evaluation of Information Access Technologies  
This paper describes the work that we conducted for the Chinese-English (CE) task of the NTCIR-10 patent machine translation evaluation.  ...  We built standard phrase-based and hierarchical phrase-based statistical machine translation (SMT) systems with optimized word segmentation, adaptive language model and improved parameter tuning strategy  ...  INTRODUCTION This paper describes our translation systems for the Chineseto-English subtask of the NTCIR-10 patent machine translation evaluation [2] .  ... 
dblp:conf/ntcir/ZhaoZUS13 fatcat:ch5ucbsgjvg5hezvg5jhaih2re

LIUM's Statistical Machine Translation System for the NTCIR Chinese/English PatentMT

Holger Schwenk, Sadaf Abdul-Rauf
2011 NTCIR Conference on Evaluation of Information Access Technologies  
This paper describes the development of a Chinese-English statistical machine translation system for the 2011 NTCIR patent translation task.  ...  Additional features include translation model adaptation using monolingual data and a continuous space language model. We report comparative results for these various configurations.  ...  We have described a new approach to adapt the translation model using monolingual data in the target language.  ... 
dblp:conf/ntcir/SchwenkA11 fatcat:k2bbf346tnd7xln32phftcvbku

The RWTH Aachen System for NTCIR-9 PatentMT

Minwei Feng, Christoph Schmidt, Joern Wuebker, Stephan Peitz, Markus Freitag, Hermann Ney
2011 NTCIR Conference on Evaluation of Information Access Technologies  
This paper describes the statistical machine translation (SMT) systems developed by RWTH Aachen University for the Patent Translation task of the 9th NTCIR Workshop.  ...  Both phrase-based and hierarchical SMT systems were trained for the constrained Japanese-English and Chinese-English tasks.  ...  Acknowledgments This work was achieved as part of the Quaero Programme, funded by OSEO, French State agency for innovation.  ... 
dblp:conf/ntcir/FengSWPFN11 fatcat:pisqn3rdovgexjimbwbz3jdfdq

Tapta: A user-driven translation system for patent documents based on domain-aware Statistical Machine Translation

Bruno Pouliquen, Christophe Mazenc, Aldo Iorio
2011 European Association for Machine Translation Conferences/Workshops  
The trained Statistical Machine Translation (SMT) tool uses this additional information to propose more accurate translations according to the context.  ...  The tool (called 'Tapta') is trained on an extensive corpus of manually translated patents. These patents are classified, each class belonging to one of the 32 predefined domains.  ...  Special thanks to the 15 persons who participated in the two tests of Tapta and to Paul Halfpenny for his valuable proof-reading.  ... 
dblp:conf/eamt/PouliquenMI11 fatcat:66jg4jiy6bfqrm5qvgcar524g4

A Japanese-to-English Statistical Machine Translation System for Technical Documents

Katsuhito Sudoh
The system achieved the BLEU scores of 34.77% and 35.75% for the NTCIR-9 and NTCIR-10 PatentMT test sets, which were consistently higher than the performance of the baseline systems using the standard  ...  This thesis addresses a Japanese-to-English statistical machine translation (SMT) system for technical documents. Machine translation (MT) is a promising solution for growing translation needs.  ...  Language Resources Japanese-to-English patent translation dataset used in NTCIR-9 Goto et al. (2011) and NTCIR-10 Goto et al. (2013) PatentMT were used for the system.  ... 
doi:10.14989/doctor.k18700 fatcat:64gd6jxxrrhfnbaj6atxjzdala

Neural machine translation - how machines learn to translate patent language

Christian Lang
2020 unpublished
Using the example of patent translation, the thesis aims to both demystify the terms "AI" and "deep-learning", that are often associated with NMT, and aims to provide an accessible guide for translators  ...  This work strives to be an easy to understand overview of how the current state-of-the-art in machine translation (MT), neural machine translation (NMT), works.  ...  Luckily MT research is still very active in Japan and the NTCIR-10 PatentMT (Patent Machine Translation) Test Collection 70 was published for research use by the National Institute of Informatics (NII)  ... 
doi:10.25365/thesis.63592 fatcat:frtadskrubajjh65r3rs2sdsna

Automated creation of domain-specific bilingual corpora for machine translation, focusing on dissimilar language pairs

Bartholomäus Wloka
2020 unpublished
The significance of sentence-aligned bilingual corpora, so-called parallel corpora, as training sets for machine translation systems and for various other language technology applications has become more  ...  The research questions addressed in this work are: How much of the text on Wikipedia content can be used to build a bilingual aligned corpus for a spe- cific language pair, and how can these texts be selected  ...  Another contribution lies in the potential use of this framework as an enhancement method for MT systems which need quick and domain-specific input of data.  ... 
doi:10.25365/thesis.65012 fatcat:nvf55v64brdl7o6fa36mrg6tcu