There is an old version of the ATIS corpus which has been compositionally
enriched with logical-semantic representations. This corpus can be
downloaded from:
ftp://mars.let.uva.nl/pub/staff/Remko_Bonnema/augmented_ATIS.tar.gz
You can find more info about this corpus in:
R. Bod, R. Bonnema and R. Scha, 1996. "A Data-Oriented Approach to
Semantic Interpretation", Proceedings Workshop on Corpus-Oriented
Semantic Analysis, ECAI-96, Budapest, Hungary. Available from cmp-lg/9606024
We also developed a much larger semantically annotated corpus (10.000
sentences), the so-called "OVIS treebank", in which each syntactic node
is enriched with a compositional semantic representation. The OVIS
treebank is used in the Dutch Priority programme Language and Speech
Technology, and will be made available very soon (hopefully this month, I
will certainly post it on this list).
More information can be found via my home page http://earth.let.uva.nl/~rens
Best,
Rens
---------------------------------------------
Rens Bod, Ph.D.
Department of Computational Linguistics
Institute for Logic, Language and Computation
Spuistraat 134, NL-1012VB Amsterdam
http://earth.let.uva.nl/~rens/
---------------------------------------------
On Tue, 7 Apr 1998, Bobby D. Bryant wrote:
> For my graduate research I need to obtain one or more corpora pairing
> transcriptions of the plaintext of sentences with some sort of
> representations of their meanings.
>
> I am aware only of ATIS3, which reportedly has at one time or another
> had SQL and/or slot-and-filler representations provided for the
> sentences. I am also aware that the LDC offers some form of ATIS3, but
> there are hints that the representations offered there are the ordinary
> sort of parse trees rather than indications of meaning. The description
> in the LDC Web pages does not give any hint as to the sort of markup
> provided, and it further indicates that whatever the markup is is far
> from complete.
>
> If anyone could elaborate on what the LDC provides for ATIS3, or direct
> me to other sources for corpora with semantic representations, I will
> appreciate it.
>
> Bobby Bryant
> Austin, Texas
>