A comparison of Statistical and Neural Machine Translation for Slovene, Serbian and Croatian

MIHAEL ARCAN

doi:10.13025/20961

A comparison of Statistical and Neural Machine Translation for Slovene, Serbian and Croatian

MIHAEL ARCAN

Computer Science

Research output: Chapter in Book or Conference Publication/Proceeding › Conference Publication › peer-review

Abstract

In this paper we present a comparison of translation quality using of Statistical Machine Translation (SMT) and Neural Machine Translation (NMT), considering translation directions between English, Slovene, Serbian and Croatian. Our experiments show that on a reduced training dataset with around two million sentences, SMT outperforms the NMT neural models. Furthermore, we present experiments with enlarged neural architectures, using 1,000 nodes and 4 hidden layers, which shows improved translation quality in terms of the BLEU metric

Original language	English (Ireland)
Title of host publication	Language Technologies and Digital Humanities 2018
Publisher	Language Technologies and Digital Humanities 2018
DOIs	https://doi.org/10.13025/20961
Publication status	Published - 1 Sept 2018

Access to Document

10.13025/20961Licence: CC BY-NC-ND

Xample-enAccepted author manuscript, 184 KBLicence: CC BY-NC-ND

Cite this

@inproceedings{477502fd85c14117b5bd460cfc367925,

title = "A comparison of Statistical and Neural Machine Translation for Slovene, Serbian and Croatian",

abstract = "In this paper we present a comparison of translation quality using of Statistical Machine Translation (SMT) and Neural Machine Translation (NMT), considering translation directions between English, Slovene, Serbian and Croatian. Our experiments show that on a reduced training dataset with around two million sentences, SMT outperforms the NMT neural models. Furthermore, we present experiments with enlarged neural architectures, using 1,000 nodes and 4 hidden layers, which shows improved translation quality in terms of the BLEU metric",

author = "MIHAEL ARCAN",

year = "2018",

month = sep,

day = "1",

doi = "10.13025/20961",

language = "English (Ireland)",

booktitle = "Language Technologies and Digital Humanities 2018",

publisher = "Language Technologies and Digital Humanities 2018",

}

TY - GEN

T1 - A comparison of Statistical and Neural Machine Translation for Slovene, Serbian and Croatian

AU - ARCAN, MIHAEL

PY - 2018/9/1

Y1 - 2018/9/1

N2 - In this paper we present a comparison of translation quality using of Statistical Machine Translation (SMT) and Neural Machine Translation (NMT), considering translation directions between English, Slovene, Serbian and Croatian. Our experiments show that on a reduced training dataset with around two million sentences, SMT outperforms the NMT neural models. Furthermore, we present experiments with enlarged neural architectures, using 1,000 nodes and 4 hidden layers, which shows improved translation quality in terms of the BLEU metric

AB - In this paper we present a comparison of translation quality using of Statistical Machine Translation (SMT) and Neural Machine Translation (NMT), considering translation directions between English, Slovene, Serbian and Croatian. Our experiments show that on a reduced training dataset with around two million sentences, SMT outperforms the NMT neural models. Furthermore, we present experiments with enlarged neural architectures, using 1,000 nodes and 4 hidden layers, which shows improved translation quality in terms of the BLEU metric

UR - http://hdl.handle.net/10379/14880

U2 - 10.13025/20961

DO - 10.13025/20961

M3 - Conference Publication

BT - Language Technologies and Digital Humanities 2018

PB - Language Technologies and Digital Humanities 2018

ER -

A comparison of Statistical and Neural Machine Translation for Slovene, Serbian and Croatian

Abstract

Access to Document

Other files and links

Fingerprint

Cite this