Background

As people at NTT and the Kyoto University Open Laboratory (i.e. members of the Hinoki project) start experimentation with the LOGON machine translation infrastructure, it will likely be useful to accumulate documentation and exchange knowledge in a piecemeal fashion. For the time being, the LOGON technology has not been released publically available, even though it will be unrestricted open-source. Thus, Hinoki members act much like beta testers, and we will keep these pages private for the time being.

Schedule and Responsibilities

• LOGON access policy and mode of distribution (oe)
• MRS property mappings (oe)
• MT bach processing and BLEU scoring (intro: oe, use: fcb)
• SEM-I interactions (use of hierarchy, look up information, et al.)
• JaEn, EnJa, NoJa, JaNo and JaJa prototypes (oe & fcb)
• sharing types across transfer grammars (oe, testing fcb)
• draft documention LogonTransfer (fcb, should really be oe)

Not strictly MT

• unknown words and numbers
• gold JaCY profiles in CVS DONE
• web demonstrator

Last update: 2008-12-01 by StephanOepen [edit]