<html>
<head>
<meta content="text/html; charset=utf-8" http-equiv="Content-Type">
</head>
<body bgcolor="#FFFFFF" text="#000000">
Hi Mike,<br>
<br>
I'm thinking about ICONS and DMRS since we anyway need that with the
MRS->DMRS conversion. The coindexed dropped arguments seems to
me fixable along the lines you suggest, but again, it's something we
need to look at for the MRS->DMRS conversion. The comment about
making sure we could express everything we needed to was more
directed at the need to find out whether there's anything
problematic when one is constructing DMRSs directly. So it would be
great if someone would suggest a suitable grammar to experiment
with, before I just decide to use the ERG ...<br>
<br>
All best,<br>
<br>
Ann<br>
<br>
<div class="moz-cite-prefix">On 05/01/2016 21:25, Michael Wayne
Goodman wrote:<br>
</div>
<blockquote
cite="mid:CAGXBFArEqUXsSSZ=HK_EiR2vVueK9srBEo_qXQvhENJzODkPkQ@mail.gmail.com"
type="cite">
<div dir="ltr">Hi Ann,
<div><br>
</div>
<div>Thanks for sharing. I couldn't find the grammar at first
because I was looking in the LOGON tree instead of the
separate LKB repository. If others are searching, it's here: <a
moz-do-not-send="true"
href="http://svn.delph-in.net/lkb/trunk/src/data/dmrscomp/"
target="_blank"><a class="moz-txt-link-freetext" href="http://svn.delph-in.net/lkb/trunk/src/data/dmrscomp/">http://svn.delph-in.net/lkb/trunk/src/data/dmrscomp/</a></a>.</div>
<div><br>
</div>
<div>I find DMRS more intuitive and more manageable than other
*MRS representations, so it's exciting to imagine a world
where that is the primary representation output by our
grammars. I'm curious to see how this works out with some
larger grammars, but I can think of a couple of challenges
(based on my discussion in Singapore: <a
moz-do-not-send="true"
href="http://moin.delph-in.net/SingaporeMrsWellformedness"
target="_blank"><a class="moz-txt-link-freetext" href="http://moin.delph-in.net/SingaporeMrsWellformedness">http://moin.delph-in.net/SingaporeMrsWellformedness</a></a>).</div>
<div><br>
</div>
<div>1. We don't yet have a way to represent ICONS in DMRS</div>
<div><br>
</div>
<div>2. DMRS currently can't express coindexed dropped arguments
(where in MRS the 'i' variable of two arguments is the same;
perhaps this can be represented using ICONS instead, or by
(re)introducing zero-pronouns)</div>
<div><br>
</div>
<div>These are both difficulties with the resulting
representation. I'm not sure if there are other issues when
implemented in the grammar. Sometime soon it would be good to
iron out these representational wrinkles. Considering ICONS, I
don't think we can just put a post-post-slash label on a link
(e.g. ARG1/NEQ/topic) because I don't think ICONS follow
normal dependency relations (Sanghoun could confirm).</div>
<div dir="ltr">
<div><br>
</div>
<br>
<div class="gmail_quote">
<div dir="ltr">On Fri, Dec 18, 2015 at 9:46 AM Ann Copestake
<<a moz-do-not-send="true"
href="mailto:aac10@cl.cam.ac.uk" target="_blank">aac10@cl.cam.ac.uk</a>>
wrote:<br>
</div>
<blockquote class="gmail_quote" style="margin:0 0 0
.8ex;border-left:1px #ccc solid;padding-left:1ex">I have
just checked in to the LKB svn repo a small grammar -
dmrscomp -<br>
and some code that extracts simple DMRSs directly from the
feature<br>
structures produced by that grammar rather than going via
MRS and RMRS.<br>
This is based on the mrscomp grammar (though with some
clean up and<br>
minor extension) - there's a fairly detailed README file.
There are a<br>
fair number of items on the TO-DO list - possibly the most<br>
time-consuming one would be to make the generator code
work with this<br>
grammar, not because there's any big problem (that I can
think of) but<br>
because the generator is quite complicated. There is also
a promise of<br>
more detailed notes, which I will supply relatively soon,
I hope - this<br>
was an interesting exercise in thinking through semantic
composition.<br>
<br>
If someone would like to collaborate on trying a similar
exercise with a<br>
larger grammar, I'd be very interested. It would help if
it were a<br>
grammar which already had the characteristic variable
property, in which<br>
case I think the main part of the conversion should be
fairly easy.<br>
<br>
There are a number of potential advantages in constructing
DMRS<br>
directly, including the ability to construct a DMRS forest
directly from<br>
a parse forest. I would argue that it also enforces some
notions of<br>
semantic well-formedness more directly than is possible
with MRS -<br>
obviously including the (equivalent of) characteristic
variable<br>
property. The semantic `fingerprint' of constructions can
be expressed<br>
more simply, because DMRS removes much of the redundancy
of MRS. But,<br>
of course, this is only interesting if we really can
express everything<br>
we want to with DMRS.<br>
<br>
All best,<br>
<br>
Ann<br>
<br>
<br>
<br>
</blockquote>
</div>
</div>
</div>
</blockquote>
<br>
</body>
</html>