[developers] New ERG with improved tokenization/preprocessing for PET

Francis Bond fcbond at gmail.com
Mon May 18 15:14:35 CEST 2009


Sorry for the slow response.

It appears that this grammar will not generate unknown words, although
I for one had been hoping it would.

 (mt::parse-interactively "Frodo barks.")
TSNLP(10): [22:09:43] translate(): read 1 MRS as generator input.
[22:09:43] translate(): processing MRS # 0 (6 EPs).
[22:09:43] translate(): error `invalid predicates: |named_unk_rel("Frodo")|'.

This is with terg+tnt (with option -mrs) as the interactive cpu, and
terg running as the top level grammar.  Batch processing has the same

Did unknown word generation not make into the mainstream?  If so, is
there a branch that has it?

Francis Bond <http://www2.nict.go.jp/x/x161/en/member/bond/>
NICT Language Infrastructure Group

More information about the developers mailing list