Can Out-of-the-box NMT Beat a Domain-trained Moses on Technical Data?. Beyer, A., Macketanz, V., Burchardt, A., & Williams, P. In Proceedings for EAMT 2017 User Studies and Project/Product Descriptions, pages 41–46, Prague, Czech Republic, 2017. ![Can Out-of-the-box NMT Beat a Domain-trained Moses on Technical Data? [pdf]](https://bibbase.org/img/filetypes/pdf.svg "pdf")
Paper abstract bibtex 4 downloads In the last year, we have seen a lot of evidence about the superiority of neural machine translation approaches (NMT) over phrase-based statistical approaches (PBMT). This trend has shown for the general domain at public competitions such as the WMT challenges as well as in the obvious quality increase in online translation services that have changed their technology. In this paper, we take the perspective of an LSP. The questions we want to answer with this study is if now is already the time to invest in the new technology. To answer this question, we have collected evidence as to whether an existing stateof-the-art NMT system for the general domain can already compete with a domaintrained and optimised Moses (PBMT) system or if it is maybe already better. As it is well known that automatic quality measures are not reliable for comparing the performance of different system types, we have performed a detailed manual evaluation based on a test suite of domain segments.
@inproceedings{beyer_can_2017,
address = {Prague, Czech Republic},
title = {Can {Out}-of-the-box {NMT} {Beat} a {Domain}-trained {Moses} on {Technical} {Data}?},
url_Paper = {https://ufal.mff.cuni.cz/eamt2017/user-project-product-papers/papers/user/EAMT2017_paper_32.pdf},
abstract = {In the last year, we have seen a lot of evidence about the superiority of neural machine translation approaches (NMT) over phrase-based statistical approaches (PBMT). This trend has shown for the general domain at public competitions such as the WMT challenges as well as in the obvious quality increase in online translation services that have changed their technology. In this paper, we take the perspective of an LSP. The questions we want to answer with this study is if now is already the time to invest in the new technology. To answer this question, we have collected evidence as to whether an existing stateof-the-art NMT system for the general domain can already compete with a domaintrained and optimised Moses (PBMT) system or if it is maybe already better. As it is well known that automatic quality measures are not reliable for comparing the performance of different system types, we have performed a detailed manual evaluation based on a test suite of domain segments.},
language = {en},
booktitle = {Proceedings for {EAMT} 2017 {User} {Studies} and {Project}/{Product} {Descriptions}},
author = {Beyer, Anne and Macketanz, Vivien and Burchardt, Aljoscha and Williams, Philip},
year = {2017},
pages = {41--46}
}
Downloads: 4
{"_id":"xF8HF6PP8gQ9vYhxr","bibbaseid":"beyer-macketanz-burchardt-williams-canoutoftheboxnmtbeatadomaintrainedmosesontechnicaldata-2017","authorIDs":["2tDjTDCzR5tC9XAza","73e5cdroyonB5H8PR","9wJCHzmZhfxTqSKLh","CyRaHJ6GyiJNprzzY","M9bqb4ELF4aKfJDDt","Me4ZHxHYxuh2zrJWJ","QXzTb6PLTiypuYXjP","S5SYCKXyx835Dktgt","WytNaKoDsAcuT6iAv","aN7JyZw5FP8ZdFxwL","ePvhWPsYksMM7Z3ki","fezDGD8KmR7542MgC","hX2SEtQeCmneRfXb8","m4pqFxB6q2rdSMMRA","mNdpXEW5TmJf74Xtq","tzJwR2nH6PPPYbF4T","zrKJ48P9qAtHpXS5J"],"author_short":["Beyer, A.","Macketanz, V.","Burchardt, A.","Williams, P."],"bibdata":{"bibtype":"inproceedings","type":"inproceedings","address":"Prague, Czech Republic","title":"Can Out-of-the-box NMT Beat a Domain-trained Moses on Technical Data?","url_paper":"https://ufal.mff.cuni.cz/eamt2017/user-project-product-papers/papers/user/EAMT2017_paper_32.pdf","abstract":"In the last year, we have seen a lot of evidence about the superiority of neural machine translation approaches (NMT) over phrase-based statistical approaches (PBMT). This trend has shown for the general domain at public competitions such as the WMT challenges as well as in the obvious quality increase in online translation services that have changed their technology. In this paper, we take the perspective of an LSP. The questions we want to answer with this study is if now is already the time to invest in the new technology. To answer this question, we have collected evidence as to whether an existing stateof-the-art NMT system for the general domain can already compete with a domaintrained and optimised Moses (PBMT) system or if it is maybe already better. As it is well known that automatic quality measures are not reliable for comparing the performance of different system types, we have performed a detailed manual evaluation based on a test suite of domain segments.","language":"en","booktitle":"Proceedings for EAMT 2017 User Studies and Project/Product Descriptions","author":[{"propositions":[],"lastnames":["Beyer"],"firstnames":["Anne"],"suffixes":[]},{"propositions":[],"lastnames":["Macketanz"],"firstnames":["Vivien"],"suffixes":[]},{"propositions":[],"lastnames":["Burchardt"],"firstnames":["Aljoscha"],"suffixes":[]},{"propositions":[],"lastnames":["Williams"],"firstnames":["Philip"],"suffixes":[]}],"year":"2017","pages":"41–46","bibtex":"@inproceedings{beyer_can_2017,\n address = {Prague, Czech Republic},\n title = {Can {Out}-of-the-box {NMT} {Beat} a {Domain}-trained {Moses} on {Technical} {Data}?},\n url_Paper = {https://ufal.mff.cuni.cz/eamt2017/user-project-product-papers/papers/user/EAMT2017_paper_32.pdf},\n abstract = {In the last year, we have seen a lot of evidence about the superiority of neural machine translation approaches (NMT) over phrase-based statistical approaches (PBMT). This trend has shown for the general domain at public competitions such as the WMT challenges as well as in the obvious quality increase in online translation services that have changed their technology. In this paper, we take the perspective of an LSP. The questions we want to answer with this study is if now is already the time to invest in the new technology. To answer this question, we have collected evidence as to whether an existing stateof-the-art NMT system for the general domain can already compete with a domaintrained and optimised Moses (PBMT) system or if it is maybe already better. As it is well known that automatic quality measures are not reliable for comparing the performance of different system types, we have performed a detailed manual evaluation based on a test suite of domain segments.},\n language = {en},\n booktitle = {Proceedings for {EAMT} 2017 {User} {Studies} and {Project}/{Product} {Descriptions}},\n author = {Beyer, Anne and Macketanz, Vivien and Burchardt, Aljoscha and Williams, Philip},\n year = {2017},\n pages = {41--46}\n}\n\n","author_short":["Beyer, A.","Macketanz, V.","Burchardt, A.","Williams, P."],"key":"beyer_can_2017","id":"beyer_can_2017","bibbaseid":"beyer-macketanz-burchardt-williams-canoutoftheboxnmtbeatadomaintrainedmosesontechnicaldata-2017","role":"author","urls":{" paper":"https://ufal.mff.cuni.cz/eamt2017/user-project-product-papers/papers/user/EAMT2017_paper_32.pdf"},"metadata":{"authorlinks":{"beyer, a":"https://annebeyer.github.io/publications/"}},"downloads":4},"bibtype":"inproceedings","biburl":"AnneBeyer.github.io/assets/bib/mine.bib","creationDate":"2020-05-05T14:13:41.447Z","downloads":4,"keywords":[],"search_terms":["out","box","nmt","beat","domain","trained","moses","technical","data","beyer","macketanz","burchardt","williams"],"title":"Can Out-of-the-box NMT Beat a Domain-trained Moses on Technical Data?","year":2017,"dataSources":["SimrqKcZkL8khGgx2","vJAkHsXDgEhvMakMj","2HmTC5wBuT2dJ6P9y","mhE92HQ6CmLusyAAc","MxT4RFGdqC49ntAWf"]}