Patent Translation Task at NTCIR Workshops

5 May 2010 13:50

The NTCIR Workshop is devoted to the Evaluation of Information Access Technologies for Asian languages and in the last two editions is including a patent translation task:

PAT-MT: Patent Translation

  • Translation Subtask
  • Cross-Lingual Information Retrieval Subtask
  • Evaluation Subtask

NTCIR PAT-MT provides a large-scale English-Japanese sentence alignment data (1.8M sentence-pairs or more. The world largest). The importance of MT has been increased in CLIR -- not only for query translation, but also for bi-directional approach in which both documents and queries are translated. There are lots of space to be investigated in MT. -- so let's work together.

Japanese it outside the scope of MOLTO, but maybe some components of our system could be tested there. Anyway, we have time to think about it, next call won't be until the beginning of 2011 at least.

More information about the workshop can be found in the official web page (