--- title: "Eurotra, history and results — Maegaard 1995" description: "The paper summarises the EUROTRA programme of the European Community (1982 to 1992), which aimed at a multilingual translation prototype for all official…" authors: ["Bente Maegaard"] year: 1995 venue: "Proceedings of Machine Translation Summit V" source: "https://aclanthology.org/1995.mtsummit-1.5/" licence: "CC BY-NC-SA 3.0 (ACL Anthology, before 2016)" licence_url: "https://creativecommons.org/licenses/by-nc-sa/3.0/" summary_licence: "Elamur Cosmic License" canonical: "https://research.elamur.ai/free/en/papers/maegaard-1995-eurotra-history-results/" --- [Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/maegaard-1995-eurotra-history-results/) [home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [1995](https://md.research.elamur.ai/free/en/papers/1995/) / Eurotra, history and results # Eurotra, history and results **Authors:** [Bente Maegaard](https://md.research.elamur.ai/free/en/authors/bente-maegaard/) · **Year:** 1995 · **Venue:** Proceedings of Machine Translation Summit V **Link:** [aclanthology.org/1995.mtsummit-1.5](https://aclanthology.org/1995.mtsummit-1.5/) · **Licence:** [CC BY-NC-SA 3.0](https://creativecommons.org/licenses/by-nc-sa/3.0/) — summary only [Proceedings of Machine Translation Summit V](https://aclanthology.org/volumes/1995.mtsummit-1/) · [licence](https://creativecommons.org/licenses/by-nc-sa/3.0/) - [Publisher's page](https://aclanthology.org/1995.mtsummit-1.5/) - [PDF](https://aclanthology.org/1995.mtsummit-1.5.pdf) **Authors (1)** 1. Maegaard, Bente [Semantic Scholar](https://www.semanticscholar.org/author/2021893) Citations and author metrics: OpenAlex, ; authors with the same name are not always told apart. ## What it is about The paper summarises the EUROTRA programme of the European Community (1982 to 1992), which aimed at a multilingual translation prototype for all official languages and drew on more than two hundred researchers at its peak. The programme left a prototype, large grammars and dictionaries for most languages, and term dictionaries. The main result it names is PaTrans, a production system built on EUROTRA technology that translates chemical patent texts from English to Danish, with a vocabulary of about 6,000 general entries and customer-maintained term bases. Two software changes made it work in practice: a faster parser and a failsoft mechanism that outputs partial structures for ungrammatical or unknown input and marks those spans for the post-editor. The author reports that the system saves 60 to 75 percent of the translator costs in use. ## Ideas - I1. Failsoft marking in the encoder: when a meta token has no chunk in any context, emit the bare form and mark its span, as the failsoft brackets do. Then count how often the marked spans are wrong, for each author pair. - I2. A layered term base: a domain term list consulted before the general profile, with an explicit priority order. In Mova this is an author or style delta over the standard, and a sense that differs by field (the paper's chemistry against music example) is kept apart by the layer, not by a special word. - I3. A post-edit measure next to chrF: sample verses, count the tokens a reader would change, and use that count as a second score. It is cheap, and it shows where exact match hides small faults. ## For Mova The code is `tools/translate/src/profile.rs` (the encoder and its back-off chain) and the outputs of `tools/roundtrip`. The paper is a history and a production report, not a method, so its use is in the failsoft and layering ideas; both are measured, not claimed. Topics: [Machine translation](https://md.research.elamur.ai/free/en/topics/machine-translation/) ## Cite The paper itself, not this card. BibTeX from the library's record of the work. ```bibtex @inproceedings{maegaard1995eurotra, title = {{Eurotra, history and results}}, author = {Maegaard, Bente}, booktitle = {Proceedings of Machine Translation Summit V}, year = {1995}, url = {https://aclanthology.org/1995.mtsummit-1.5/} } ``` - [← previous Corpus Statistics Meet the Noun Compound: Some Empirical Results](https://md.research.elamur.ai/free/en/papers/aclbulk-1995-lauer-corpus-statistics-meet-noun-compound/) - [next → Evaluating Automated and Manual Acquisition of Anaphora Resolution Strategies](https://md.research.elamur.ai/free/en/papers/aone-1995-evaluating-automated-manual-acquisition-anaphora/) - [↑ back to papers of 1995](https://md.research.elamur.ai/free/en/papers/1995/) ## References 1 references, 0 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet. 1. The EUROTRA technical specifications and formal specifications are presented in Copeland, Durand, Krauwer, Maegaard (ed.): Studies in Natural Language Processing, Vol. 1 and 2, CEC, Luxembourg, 1991. [find](https://scholar.google.com/scholar?q=The+EUROTRA+technical+specifications+and+formal+specifications+are+presented+in+Copeland%2C+Durand%2C+Krauwer%2C+Maegaard+%28ed%2E%29%3A+Studies+in+Natural+Language+Processing%2C+Vol%2E+1+and+2%2C+CEC%2C+Luxembourg%2C+1991%2E)