--- title: "Roland Kuhn — author profile and papers" description: "Roland Kuhn, National Research Council Canada: 3 papers on Elamur Research (2007–2010), h-index 24 — affiliations, topics, co-authors and who cites them…" canonical: "https://research.elamur.ai/free/en/authors/roland-kuhn/" --- [Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/authors/roland-kuhn/) [home](https://md.research.elamur.ai/free/en/) / [authors](https://md.research.elamur.ai/free/en/authors/) / Roland Kuhn # Roland Kuhn Roland Kuhn works on statistical machine translation, and the papers here centre on phrase-based systems. Two questions recur in them. One is how to make a system trained on one kind of text translate well in another domain. The other is how to tell useful entries in large phrase tables from noise. The methods are statistical: mixtures of language and translation models, weights fitted by maximum likelihood or EM, discriminative classifiers, and significance tests applied to counts drawn from word-aligned parallel corpora. The most cited work here is [Mixture-Model Adaptation for SMT](https://md.research.elamur.ai/free/en/papers/foster-2007-mixture-model-adaptation-smt/) (2007). It splits a Chinese-to-English training corpus by source and trains separate component models on each part. It then mixes them according to an estimate of each part's closeness to the material being translated. The aim is to weight every component softly rather than keep some data and discard the rest. Also from 2007, [Improving Translation Quality by Discarding Most of the Phrasetable](https://md.research.elamur.ai/free/en/papers/johnson-2007-improving-translation-quality-discarding-most/) uses Fisher's exact test to ask whether each phrase pair co-occurs more often than chance would allow. Pairs that fail the test are removed, which shrinks the table without the quality loss that cruder pruning tends to cause. The newest paper here, [Discriminative Instance Weighting for Domain Adaptation](https://md.research.elamur.ai/free/en/papers/foster-2010-discriminative-instance-weighting-domain-adaptation/) (2010), returns to domain adaptation at a finer grain. Instead of weighting whole corpora or sentences, it gives each out-of-domain phrase pair its own weight. That weight combines a measure of similarity to the in-domain data with a separate estimate of how general the pair's language is, partly learned by an SVM. All the parameters are then tuned together on a small in-domain development set. Kuhn's current affiliation is the [National Research Council Canada](https://md.research.elamur.ai/free/en/government/national-research-council-canada/) in Ottawa, dating from 2005. Earlier papers came from the Computer Research Institute of Montréal and [McGill University](https://md.research.elamur.ai/free/en/universities/mcgill-university/). The most frequent co-author here is [George Foster](https://md.research.elamur.ai/free/en/authors/george-foster/), with [Cyril Goutte](https://md.research.elamur.ai/free/en/authors/cyril-goutte/) also appearing. The papers fall under [Machine translation](https://md.research.elamur.ai/free/en/topics/machine-translation/), with one also touching [Evaluation & benchmarks](https://md.research.elamur.ai/free/en/topics/evaluation-benchmarks/). Roland Kuhn has 3 papers in our library, in 2007–2010, most of them on [Machine translation](https://md.research.elamur.ai/free/en/topics/machine-translation/), [Evaluation & benchmarks](https://md.research.elamur.ai/free/en/topics/evaluation-benchmarks/); often with [George Foster](https://md.research.elamur.ai/free/en/authors/george-foster/), [Cyril Goutte](https://md.research.elamur.ai/free/en/authors/cyril-goutte/). 3 papers in our library · 2007–2010 [OpenAlex](https://openalex.org/A5103230324) · [ORCID](https://orcid.org/0000-0002-1485-140X) · [**h-index 24**](https://openalex.org/authors/A5103230324) · [98 works](https://openalex.org/works?filter=author.id:A5103230324) · [3,333 citations](https://openalex.org/works?filter=author.id:A5103230324&sort=cited_by_count:desc) OpenAlex, 2026-10-14 [📄 3 works in our library](https://md.research.elamur.ai/free/en/authors/roland-kuhn/#papers) · [🔗 544 citations in our library](https://md.research.elamur.ai/free/en/authors/roland-kuhn/cited-by/) · [11 more in our library without a page yet](#also) · [⭐ Paper Star rank #1992](https://md.research.elamur.ai/free/en/paper-stars/page/20/#star-1992) The profile is the library's match of the name; two people with one name are not always told apart. Elsewhere: [Google Scholar](https://scholar.google.com/citations?view_op=search_authors&mauthors=Roland+Kuhn) · [Semantic Scholar](https://www.semanticscholar.org/search?q=Roland+Kuhn) · [dblp](https://dblp.org/search?q=Roland+Kuhn) · [ACL Anthology](https://aclanthology.org/search/?q=Roland+Kuhn) ## Affiliations [🏛️**National Research Council Canada** Ottawa · Canada 2005–2025 · 30 papers*current*](https://md.research.elamur.ai/free/en/government/national-research-council-canada/) · [🏛️**National Academies of Sciences, Engineering, and Medicine** Washington · United States 2010–2020 · 8 papers](https://md.research.elamur.ai/free/en/government/national-academies-of-sciences-engineering-and-medicine/) · [🤝**Computer Research Institute of Montréal** Montreal · Canada 1993–2002 · 8 papers](https://ror.org/0279d5115) · [🎓**McGill University** Montreal · Canada 1988–2002 · 7 papers](https://md.research.elamur.ai/free/en/universities/mcgill-university/) · [🏢**Panasonic** Newark · United States 2000–2004 · 4 papers](https://ror.org/00pvgrv63) · [🎓**TU Dortmund University** Dortmund · Germany 2017–2021 · 3 papers](https://md.research.elamur.ai/free/en/universities/tu-dortmund-university/) · [🏢**Panasonic** Kadoma · Japan 2000–2004 · 3 papers](https://ror.org/011tm7n37) From the bylines of their papers (OpenAlex, 2026-10-14), ORCID and Wikidata; institutions on papers are matched by machine. 1 more institution on single bylines are not listed. ## Topics [Machine translation (3)](https://md.research.elamur.ai/free/en/topics/machine-translation/) · [Evaluation & benchmarks (1)](https://md.research.elamur.ai/free/en/topics/evaluation-benchmarks/) ## Works by year ### 2010s [**2010** 1 paper](https://md.research.elamur.ai/free/en/papers/foster-2010-discriminative-instance-weighting-domain-adaptation/) ### 2000s [**2007** 2 papers](https://md.research.elamur.ai/free/en/authors/roland-kuhn/year/2007/) ## Papers The most cited first (OpenAlex and our library). 1. **[Mixture-Model Adaptation for SMT](https://md.research.elamur.ai/free/en/papers/foster-2007-mixture-model-adaptation-smt/)** *(2007 · 397 citations)* SMT systems trained on parallel data from narrow sources, such as parliamentary proceedings, often perform poorly on other domains. 2. **[Discriminative Instance Weighting for Domain Adaptation in Statistical Machine Translation](https://md.research.elamur.ai/free/en/papers/foster-2010-discriminative-instance-weighting-domain-adaptation/)** *(2010 · 305 citations)* The paper tackles domain adaptation for phrase-based SMT. A large out-of-domain parallel corpus (OUT) is used to help a smaller in-domain one (IN), which is hard when the two differ a lot. The authors do not decide at the level of whole sentences or corpora. Instead, they assign a weight to every OUT phrase pair. The… 3. **[Improving Translation Quality by Discarding Most of the Phrasetable](https://md.research.elamur.ai/free/en/papers/johnson-2007-improving-translation-quality-discarding-most/)** *(2007 · 300 citations)* Phrase-based statistical machine translation systems extract huge phrasetables from word-aligned parallel corpora. Many of the entries look wrong or are never used, yet crude size reductions usually cost BLEU. Johnson, Martin, Foster and Kuhn (NRC Canada) propose filtering each phrase pair with a significance test. ## Also in our library 11 more works of theirs are in our library without a page here yet; they will appear as we publish them. ## Co-authors here 2 co-authors with a page, the most papers together first. - [George Foster](https://md.research.elamur.ai/free/en/authors/george-foster/) ×3 - [Cyril Goutte](https://md.research.elamur.ai/free/en/authors/cyril-goutte/) ## Cited by Papers of our library cite their works 544 times (of the 140,884 papers whose bibliographies we read); [23 citing papers with a page here →](https://md.research.elamur.ai/free/en/authors/roland-kuhn/cited-by/).