Context Adaptation in Statistical Machine Translation Using Models with Exponentially Decaying Cache

Research output: Chapter in Book/Report/Conference proceedingConference contributionScientificpeer-review

Abstract

We report results from a domain adaptation task for statistical machine translation (SMT) using cache-based adaptive language and translation models. We apply an exponential decay factor and integrate the cache models in a standard phrasebased SMT decoder. Without the need for any domain-specific resources we obtain a 2.6% relative improvement on average in BLEU scores using our dynamic adaptation procedure.
Original languageOther/Unknown
Title of host publicationUnknown host publication
Number of pages8
Publication date1 Jul 2010
Pages8-15
Publication statusPublished - 1 Jul 2010
MoE publication typeA4 Article in conference proceedings
EventUnknown host publication - , International
Duration: 1 Jan 1800 → …

Cite this

@inproceedings{75deb8c94edc49a0bc29916668b473b1,
title = "Context Adaptation in Statistical Machine Translation Using Models with Exponentially Decaying Cache",
abstract = "We report results from a domain adaptation task for statistical machine translation (SMT) using cache-based adaptive language and translation models. We apply an exponential decay factor and integrate the cache models in a standard phrasebased SMT decoder. Without the need for any domain-specific resources we obtain a 2.6{\%} relative improvement on average in BLEU scores using our dynamic adaptation procedure.",
author = "J{\"o}rg Tiedemann",
note = "Volume: Proceeding volume:",
year = "2010",
month = "7",
day = "1",
language = "odefinierat/ok{\"a}nt",
pages = "8--15",
booktitle = "Unknown host publication",

}

Tiedemann, J 2010, Context Adaptation in Statistical Machine Translation Using Models with Exponentially Decaying Cache. in Unknown host publication. pp. 8-15, Unknown host publication, International, 01/01/1800.

Context Adaptation in Statistical Machine Translation Using Models with Exponentially Decaying Cache. / Tiedemann, Jörg.

Unknown host publication. 2010. p. 8-15.

Research output: Chapter in Book/Report/Conference proceedingConference contributionScientificpeer-review

TY - GEN

T1 - Context Adaptation in Statistical Machine Translation Using Models with Exponentially Decaying Cache

AU - Tiedemann, Jörg

N1 - Volume: Proceeding volume:

PY - 2010/7/1

Y1 - 2010/7/1

N2 - We report results from a domain adaptation task for statistical machine translation (SMT) using cache-based adaptive language and translation models. We apply an exponential decay factor and integrate the cache models in a standard phrasebased SMT decoder. Without the need for any domain-specific resources we obtain a 2.6% relative improvement on average in BLEU scores using our dynamic adaptation procedure.

AB - We report results from a domain adaptation task for statistical machine translation (SMT) using cache-based adaptive language and translation models. We apply an exponential decay factor and integrate the cache models in a standard phrasebased SMT decoder. Without the need for any domain-specific resources we obtain a 2.6% relative improvement on average in BLEU scores using our dynamic adaptation procedure.

M3 - Konferensbidrag

SP - 8

EP - 15

BT - Unknown host publication

ER -