Philipp Koehn

Philipp Koehn is a Professor in the Department of Computer Science at Johns Hopkins University, with additional affiliation at the University of Edinburgh. His primary research focuses on statistical and neural machine translation, specifically developing methods to leverage large-scale digital information for cross-lingual communication. He leads the Machine Translation Research Group and maintains key resources like the Moses toolkit and Europarl corpus.

His research interests span:

• Core machine translation techniques (statistical/neural approaches)
• Low-resource and unsupervised translation methods
• Cross-lingual representation learning
• Speech-to-speech translation systems
• Large-scale parallel data mining and alignment
• Evaluation methodologies for generated text

Koehn's recent publications demonstrate strong focus on improving translation efficiency (dynamic compression, streaming models), robustness (noise handling, error correction), and accessibility (low-resource languages, radio speech processing). Key trends include multilingual generalization, document-level coherence, and human-centered evaluation.

Significant scientific recognition includes:

• ACL Fellow (2024)
• IAMT Award of Honor (2015)
• European Inventor Award Finalist (2013)

He currently advises PhD students Rachel Wicks, Elina Baral, Bismarck Odoom, and Weiting Tan. His Machine Translation Group develops widely-used open-source tools and organizes major conferences including WMT and MT Marathon.