Rico Sennrich

Dr. Rico Sennrich
Lecturer (Assistant Professor)
School of Informatics, room 3.03A
10 Crichton Street
Edinburgh, EH8 9AB, United Kingdom
rico.sennrich AT ed.ac.uk

I am a Lecturer (Assistant Professor) at the University of Edinburgh working on machine learning, especially in the area of machine translation and natural language processing. My recent focus has been high-quality machine translation, with syntax-based SMT, and with neural machine translation. Some of our results can be found in my publications, and in the results of the WMT shared translation tasks (2014, 2015, 2016, 2017).
I am a member of the Machine Translation Group and the Edinburgh NLP Group - come join our weekly meetings.
I am also affiliated with the University of Zurich in the CoNTra project.


  • 2019: congratulations to Elena Voita, Biao Zhang, and all co-authors: 4 papers accepted to ACL 2019.
  • 2019: I have an open postdoc researcher position at the University of Zurich (now filled).
  • 2018: I have two funded PhD studentships at the University of Zurich (start: September 2019; application deadline: 20 January 2019).
  • 2018: I received a SNSF Professorship and will move to the University of Zurich in August 2019.
  • 2018: congratulations to Jonathan Mallinson, Gongbo Tang, Samuel Läubli, and all co-authors: 3 papers accepted to EMNLP 2018.
  • 2018: congratulations to Elena Voita; her paper on anaphora resolution in discourse-level MT has been accepted to ACL.
  • 2018: congratulations to Rachel Bawden; her paper on discourse-level MT has been accepted to NAACL.
  • 2017: I have been appointed Lecturer in Machine Learning at the University of Edinburgh.
  • 2017: congratulation to my co-authors: 2 short papers accepted to EMNLP; 3 research papers to WMT. Camera-ready versions coming soon.
  • 2017: UEDIN is tied for best constrained system for 11 out of 12 translation pairs in which we participated in the WMT17 shared news translation task.
  • 2016: how grammatical is character-level neural MT? Difficult question, but we can test this with contrastive translation pairs (EACL 2017). The test set is released here.
  • 2016: update: now all our WMT16 neural systems are available here.
  • 2016: new textbook: Philip Williams, Rico Sennrich, Matt Post and Philipp Koehn. Syntax-based Statistical Machine Translation
  • 2016: Edinburgh submitted neural MT systems to the WMT 16 shared translation task for 8 translation directions, which are ranked (tied) best for 7 out of 8. Get code and data here. System description here

Prospective interns and visitors

I do not currently have any openings for internships. Please do not send me an unsolicited application.