539 research outputs found
Sort by
Dictionary-based Domain Adaptation of MT Systems without Retraining
Get PDFWe describe our submission to the IT-domain translation task of WMT 2016.
We perform domain adaptation with dictionary data on already trained MT systems with no further retraining.
We apply our approach to two conceptually different systems developed within the QTLeap project: TectoMT and Moses, as well as Chimera, their combination.
In all settings, our method improves the translation quality.
Moreover, the basic variant of our approach is applicable to any MT system, including a black-box one
CzEng 1.6
No full textA new release of the Czech-English parallel corpus CzEng. CzEng 1.6 consists of about 0.5 billion words (“gigaword”) in each language. The corpus is equipped with automatic annotation at a deep syntactic level of representation and alternatively in Universal Dependencies
Moses & Treex Hybrid MT Systems Bestiary
No full textMoses is a well-known representative of the phrase-based statistical machine translation systems family, which are known to be extremely poor in explicit linguistic knowledge, operating on flat language representations, consisting only of tokens and phrases. Treex, on the other hand, is a highly linguistically motivated NLP toolkit, operating on several layers of language representation, rich in linguistic annotations. Its main application is TectoMT, a hybrid machine translation system with deep syntax transfer. We review a large number of machine translation systems that have been built over the past years by combining Moses and Treex/TectoMT in various ways
Merged bilingual trees based on Universal Dependencies in Machine Translation
Get PDFIn this paper, we present our new experimental system of merging dependency
representations of two parallel sentences
into one dependency tree. All the inner nodes in dependency tree represent source-target pairs of words, the extra words are in form of leaf nodes. We use Universal Dependencies annotation style, in which the function words, whose usage often differs between languages, are annotated as leaves.
The parallel treebank is parsed in minimally supervised way. Unaligned words are there automatically pushed to leaves. We present a simple translation system trained on such merged trees and evaluate it in WMT 2016 English-to-Czech and Czech-to-English translation task. Even though the model is so far very simple and no language model and word-reordering model were used, the Czech-to-English variant reached similar BLEU score as another established tree-based system
WMT Metrics Task: Past Great Metrics, Current Evaluation Experiments, Future Metrics Challenges
No full textA summary of ten years of automatic MT evaluation, highlighting past successful metrics, current evaluation challenges and topics that need to be tackled in the future
Results of the WMT16 Tuning Shared Task
Get PDFThis paper presents the results of the
WMT16 Tuning Shared Task. We provided
the participants of this task with a
complete machine translation system and
asked them to tune its internal parameters
(feature weights). The tuned systems were
used to translate the test set and the outputs
were manually ranked for translation
quality. We received 4 submissions in the
Czech-English and 8 in the English-Czech
translation direction. In addition, we ran
2 baseline setups, tuning the parameters
with standard optimizers for BLEU score.
In contrast to previous years, the tuned
systems in 2016 rely on large data
The QT21/HimL Combined Machine Translation System
Get PDFThis paper describes the joint submission
of the QT21 and HimL projects for
the English→Romanian translation task of
the ACL 2016 First Conference on Machine
Translation (WMT 2016). The submission
is a system combination which
combines twelve different statistical machine
translation systems provided by the
different groups (RWTH Aachen University,
LMU Munich, Charles University in
Prague, University of Edinburgh, University
of Sheffield, Karlsruhe Institute of
Technology, LIMSI, University of Amsterdam,
Tilde). The systems are combined
using RWTH’s system combination
approach. The final submission shows an
improvement of 1.0 BLEU compared to the
best single system on newstest2016
Reranking Hypotheses of Machine-Translated Queries for Cross-Lingual Information Retrieval
No full textMachine Translation (MT) systems employed to translate queries for Cross-Lingual Information Retrieval typically produce single translation with maximum translation quality. This, however, might not be optimal with respect to retrieval quality and other translation variants might lead to better retrieval results. In this paper, we explore a method exploiting multiple translations produced by an MT system, which are reranked using a supervised machine-learning method trained to directly optimize the retrieval quality. We experiment with various types of features and the results obtained on the medical-domain test collection from the CLEF eHealth Lab series show significant improvement of retrieval quality compared to a system using single translation provided by MT
CzEng 1.6 pipeline: Docker container
No full textA pipeline of linguistic analysis used for CzEng 1.6 packaged in a Docker container. Any English or Czech text can be processed by this pipeline. The processing pipeline includes part-of-speech tagging, parsing, named entity recognition, semantic role labelling, coreference resolution etc
Coreference in Prague Czech-English Dependency Treebank
No full textWe present coreference annotation on parallel Czech-English texts of the Prague Czech-English Dependency Treebank (PCEDT). The paper describes innovations made to PCEDT 2.0 concerning coreference, as well as coreference information already present there. We characterize the coreference annotation scheme, give the statistics and compare our annotation with the coreference annotation in Ontonotes and Prague Dependency Treebank for Czech. We also present the experiments made using this corpus to improve the alignment of coreferential expressions, which helps us to collect better statistics of correspondences between types of coreferential relations in Czech and English. The corpus released as PCEDT 2.0 Coref is publicly available