[developers] Paraphrase grammars and SEMI's
danf at stanford.edu
Tue May 26 03:26:11 CEST 2009
I'm in agreement with the rest of you. The internal type hierarchy is also (necessarily, I think), more subject to restructuring as the quest for ever better linguistic analysis proceeds. I think it's useful to view the paraphrase task as related to the MT task, for which the SEM-I MRS is essential.
----- Original Message -----
From: "Francis Bond" <fcbond at gmail.com>
To: crysmann at ifk.uni-bonn.de
Cc: developers at delph-in.net, "Darren Scott Appling" <darren.scott.appling at gmail.com>
Sent: Monday, May 25, 2009 4:03:08 PM GMT +01:00 Amsterdam / Berlin / Bern / Rome / Stockholm / Vienna
Subject: Re: [developers] Paraphrase grammars and SEMI's
2009/5/25 Berthold Crysmann <crysmann at ifk.uni-bonn.de>:
> On Mon, 2009-05-25 at 21:59 +0900, Francis Bond wrote:
>> we would like to solicit opinions (especially from Dan) about the
>> following question:
>> If you are making a paraphrase grammar, should it use the language
>> internal MRS, or the post SEM-I external MRS.
>> That is:
>> Internal: TEXT - ERG - MRS - EnEn - MRS - ERG - TEXT
>> External: TEXT -ERG - MRS - MRS' - EnEn - MRS' - MRS - ERG - TEXT
>> My view is that we should use the external MRS for the following reasons:
> We did some paraphrasing with GG in Checkpoint. At that time, the
> paraphraser was using the internal MRS. As a result, I had to
> cross-classify internal and external semantic semantic entities, such
> that the types in mtr.tdl were compatible with both universes. The
> result is not a beauty.
Thanks. That's another good reason.
Francis Bond <http://www2.nict.go.jp/x/x161/en/member/bond/>
NICT Language Infrastructure Group
More information about the developers