Interactive translation prediction versus conventional post-editing in practice: a study with the CasMaCat workbench

Interactive translation prediction versus conventional post-editing in practice: a study with the CasMaCat workbench. Sanchis-Trilles, G., Alabau, V., Buck, C., Carl, M., Casacuberta, F., García-Martínez, M., Germann, U., González-Rubio, J., Hill, R., Koehn, P., Leiva, L., Mesa-Lao, B., Ortiz-Martínez, D., Saint-Amand, H., Tsoukala, C., & Vidal, E. Machine Translation, 28(3-4):217–235, Springer, 12, 2014.
doi abstract bibtex

We conducted a field trial in computer-assisted professional translation to compare interactive translation prediction (ITP) against conventional post-editing (PE) of machine translation (MT) output. In contrast to the conventional PE set-up, where an MT system first produces a static translation hypothesis that is then edited by a professional (hence “post-editing”), ITP constantly updates the translation hypothesis in real time in response to user edits. Our study involved nine professional translators and four reviewers working with the web-based CasMaCat workbench. Various new interactive features aiming to assist the post-editor/translator were also tested in this trial. Our results show that even with little training, ITP can be as productive as conventional PE in terms of the total time required to produce the final translation. Moreover, translation editors working with ITP require fewer key strokes to arrive at the final version of their translation.

@article{dbc1aa2707a548a9adaf8fa37f85d3fd,
  title     = "Interactive translation prediction versus conventional post-editing in practice: a study with the CasMaCat workbench",
  abstract  = "We conducted a field trial in computer-assisted professional translation to compare interactive translation prediction (ITP) against conventional post-editing (PE) of machine translation (MT) output. In contrast to the conventional PE set-up, where an MT system first produces a static translation hypothesis that is then edited by a professional (hence “post-editing”), ITP constantly updates the translation hypothesis in real time in response to user edits. Our study involved nine professional translators and four reviewers working with the web-based CasMaCat workbench. Various new interactive features aiming to assist the post-editor/translator were also tested in this trial. Our results show that even with little training, ITP can be as productive as conventional PE in terms of the total time required to produce the final translation. Moreover, translation editors working with ITP require fewer key strokes to arrive at the final version of their translation.",
  author    = "Germ{\'a}n Sanchis-Trilles and Vicent Alabau and Christian Buck and Michael Carl and Francisco Casacuberta and Mercedes Garc{\'i}a-Mart{\'i}nez and Ulrich Germann and Jes{\'u}s Gonz{\'a}lez-Rubio and Hill, {Robin L.} and Philipp Koehn and Leiva, {Luis A.} and Bartolom{\'e} Mesa-Lao and Daniel Ortiz-Mart{\'i}nez and Herve Saint-Amand and Chara Tsoukala and Enrique Vidal",
  year      = "2014",
  month     = "12",
  day       = "1",
  doi       = "10.1007/s10590-014-9157-9",
  language  = "English",
  volume    = "28",
  pages     = "217--235",
  journal   = "Machine Translation",
  issn      = "0922-6567",
  publisher = "Springer",
  number    = "3-4",
}

Downloads: 0

{"_id":"qcxSQQP37bkMyZWRE","bibbaseid":"sanchistrilles-alabau-buck-carl-casacuberta-garciamartinez-germann-gonzlezrubio-etal-interactivetranslationpredictionversusconventionalposteditinginpracticeastudywiththecasmacatworkbench-2014","authorIDs":[],"author_short":["Sanchis-Trilles, G.","Alabau, V.","Buck, C.","Carl, M.","Casacuberta, F.","García-Martínez, M.","Germann, U.","González-Rubio, J.","Hill, R.","Koehn, P.","Leiva, L.","Mesa-Lao, B.","Ortiz-Martínez, D.","Saint-Amand, H.","Tsoukala, C.","Vidal, E."],"bibdata":{"bibtype":"article","type":"article","title":"Interactive translation prediction versus conventional post-editing in practice: a study with the CasMaCat workbench","abstract":"We conducted a field trial in computer-assisted professional translation to compare interactive translation prediction (ITP) against conventional post-editing (PE) of machine translation (MT) output. In contrast to the conventional PE set-up, where an MT system first produces a static translation hypothesis that is then edited by a professional (hence “post-editing”), ITP constantly updates the translation hypothesis in real time in response to user edits. Our study involved nine professional translators and four reviewers working with the web-based CasMaCat workbench. Various new interactive features aiming to assist the post-editor/translator were also tested in this trial. Our results show that even with little training, ITP can be as productive as conventional PE in terms of the total time required to produce the final translation. Moreover, translation editors working with ITP require fewer key strokes to arrive at the final version of their translation.","author":[{"firstnames":["Germán"],"propositions":[],"lastnames":["Sanchis-Trilles"],"suffixes":[]},{"firstnames":["Vicent"],"propositions":[],"lastnames":["Alabau"],"suffixes":[]},{"firstnames":["Christian"],"propositions":[],"lastnames":["Buck"],"suffixes":[]},{"firstnames":["Michael"],"propositions":[],"lastnames":["Carl"],"suffixes":[]},{"firstnames":["Francisco"],"propositions":[],"lastnames":["Casacuberta"],"suffixes":[]},{"firstnames":["Mercedes"],"propositions":[],"lastnames":["García-Martínez"],"suffixes":[]},{"firstnames":["Ulrich"],"propositions":[],"lastnames":["Germann"],"suffixes":[]},{"firstnames":["Jesús"],"propositions":[],"lastnames":["González-Rubio"],"suffixes":[]},{"propositions":[],"lastnames":["Hill"],"firstnames":["Robin L."],"suffixes":[]},{"firstnames":["Philipp"],"propositions":[],"lastnames":["Koehn"],"suffixes":[]},{"propositions":[],"lastnames":["Leiva"],"firstnames":["Luis A."],"suffixes":[]},{"firstnames":["Bartolomé"],"propositions":[],"lastnames":["Mesa-Lao"],"suffixes":[]},{"firstnames":["Daniel"],"propositions":[],"lastnames":["Ortiz-Martínez"],"suffixes":[]},{"firstnames":["Herve"],"propositions":[],"lastnames":["Saint-Amand"],"suffixes":[]},{"firstnames":["Chara"],"propositions":[],"lastnames":["Tsoukala"],"suffixes":[]},{"firstnames":["Enrique"],"propositions":[],"lastnames":["Vidal"],"suffixes":[]}],"year":"2014","month":"12","day":"1","doi":"10.1007/s10590-014-9157-9","language":"English","volume":"28","pages":"217–235","journal":"Machine Translation","issn":"0922-6567","publisher":"Springer","number":"3-4","bibtex":"@article{dbc1aa2707a548a9adaf8fa37f85d3fd,\n title = \"Interactive translation prediction versus conventional post-editing in practice: a study with the CasMaCat workbench\",\n abstract = \"We conducted a field trial in computer-assisted professional translation to compare interactive translation prediction (ITP) against conventional post-editing (PE) of machine translation (MT) output. In contrast to the conventional PE set-up, where an MT system first produces a static translation hypothesis that is then edited by a professional (hence “post-editing”), ITP constantly updates the translation hypothesis in real time in response to user edits. Our study involved nine professional translators and four reviewers working with the web-based CasMaCat workbench. Various new interactive features aiming to assist the post-editor/translator were also tested in this trial. Our results show that even with little training, ITP can be as productive as conventional PE in terms of the total time required to produce the final translation. Moreover, translation editors working with ITP require fewer key strokes to arrive at the final version of their translation.\",\n author = \"Germ{\\'a}n Sanchis-Trilles and Vicent Alabau and Christian Buck and Michael Carl and Francisco Casacuberta and Mercedes Garc{\\'i}a-Mart{\\'i}nez and Ulrich Germann and Jes{\\'u}s Gonz{\\'a}lez-Rubio and Hill, {Robin L.} and Philipp Koehn and Leiva, {Luis A.} and Bartolom{\\'e} Mesa-Lao and Daniel Ortiz-Mart{\\'i}nez and Herve Saint-Amand and Chara Tsoukala and Enrique Vidal\",\n year = \"2014\",\n month = \"12\",\n day = \"1\",\n doi = \"10.1007/s10590-014-9157-9\",\n language = \"English\",\n volume = \"28\",\n pages = \"217--235\",\n journal = \"Machine Translation\",\n issn = \"0922-6567\",\n publisher = \"Springer\",\n number = \"3-4\",\n}\n\n","author_short":["Sanchis-Trilles, G.","Alabau, V.","Buck, C.","Carl, M.","Casacuberta, F.","García-Martínez, M.","Germann, U.","González-Rubio, J.","Hill, R.","Koehn, P.","Leiva, L.","Mesa-Lao, B.","Ortiz-Martínez, D.","Saint-Amand, H.","Tsoukala, C.","Vidal, E."],"key":"dbc1aa2707a548a9adaf8fa37f85d3fd","id":"dbc1aa2707a548a9adaf8fa37f85d3fd","bibbaseid":"sanchistrilles-alabau-buck-carl-casacuberta-garciamartinez-germann-gonzlezrubio-etal-interactivetranslationpredictionversusconventionalposteditinginpracticeastudywiththecasmacatworkbench-2014","role":"author","urls":{},"downloads":0},"bibtype":"article","biburl":"http://www.research.ed.ac.uk/portal/en/organisations/design-informatics(035197e2-377f-4837-97f7-7e235a2bbb93)/publications.bibtex","creationDate":"2019-06-05T13:17:40.467Z","downloads":0,"keywords":[],"search_terms":["interactive","translation","prediction","versus","conventional","post","editing","practice","study","casmacat","workbench","sanchis-trilles","alabau","buck","carl","casacuberta","garcía-martínez","germann","gonzález-rubio","hill","koehn","leiva","mesa-lao","ortiz-martínez","saint-amand","tsoukala","vidal"],"title":"Interactive translation prediction versus conventional post-editing in practice: a study with the CasMaCat workbench","year":2014,"dataSources":["2frK9pzuQoZ7LkTLt"]}