NICT at WAT 2015
نویسندگان
چکیده
Translation systems of our NICT team at the 2nd Workshop on Asian Translation (WAT 2015) are described in this paper. We participated in two translation tasks: Japanese-to-English (JE) and Korean-toJapanese (KJ). A baseline phrased-based (PB) statistical machine translation (SMT) system in Moses was used. On JE translation, two pre-reordering approaches were applied: a simple reverse preordering and a dependency-based approach. On KJ translation, the processing was purely conducted on character-level. Evaluation results show that even simple approaches can improve JE and KJ PB SMT significantly. These techniques can be easily applied in practice because of the simplicity.
منابع مشابه
A Simple and Strong Baseline: NAIST-NICT Neural Machine Translation System for WAT2017 English-Japanese Translation Task
This paper describes the details about the NAIST-NICT machine translation system for WAT2017 English-Japanese Scientific Paper Translation Task. The system consists of a language-independent tokenizer and an attentional encoder-decoder style neural machine translation model. According to the official results, our system achieves higher translation accuracy than any systems submitted previous ca...
متن کاملEnsemble and Reranking: Using Multiple Models in the NICT-2 Neural Machine Translation System at WAT2017
In this paper, we describe the NICT-2 neural machine translation system evaluated at WAT2017. This system uses multiple models as an ensemble and combines models with opposite decoding directions by reranking (called bi-directional reranking). In our experimental results on small data sets, the translation quality improved when the number of models was increased to 32 in total and did not satur...
متن کاملNICT-2 Translation System for WAT2016: Applying Domain Adaptation to Phrase-based Statistical Machine Translation
This paper describes the NICT-2 translation system for the 3rd Workshop on Asian Translation. The proposed system employs a domain adaptation method based on feature augmentation. We regarded the Japan Patent Office Corpus as a mixture of four domain corpora and improved the translation quality of each domain. In addition, we incorporated language models constructed from Google n-grams as exter...
متن کاملNAVER Machine Translation System for WAT 2015
In this paper, we describe NAVER machine translation system for English to Japanese and Korean to Japanese tasks at WAT 2015. We combine the traditional SMT and neural MT in both tasks.
متن کاملWhite Paper: An Automatic Test Method with Improved Precision for Wax Appearance Temperature (WAT) of Crude Oils
One such method, Differential Scanning Calorimetry (DSC), detects WAT by measuring the difference in heat absorbed or released between a reference sample and the test sample at a given cooling or warming rate. The WAT is detected by a deviation of experimental data from the reference baseline and typically requires a significant amount of wax formation for sensible detection. Consequently, it m...
متن کامل