[Corpora-List] Semantic parsing for the robot commands corpus by integrating spatial knowledge

[John F Sowa]

On 9/7/2013 2:55 AM, Kais Dukes wrote:
> In a nutshell, I’m looking for any suggestions on how to do
> joint parsing / spatial disambiguation. Ideally not a pipeline
> approach, but something integrated. Surely there must be
> a good way to use spatial knowledge while parsing?

On 9/8/2013 8:18 PM, Chris Brew wrote:
> It'll be much easier to do the feedback from dialog context and
> real-world situation if you use a parser that is designed, from the
> outset, to integrate information from multiple levels of representation.

Yes.  Most parsers are designed for a pipeline.  They usually don't
provide a way to accommodate external information about context,
background knowledge, and pragmatics during the parsing stage.

On 9/8/2013 9:39 AM, Kilian Evang wrote:
> the first possibility that comes to my mind is using a chart parser
> where producing semantic representations is integrated with the
> syntactic parsing process. I.e. directly produce the semantics of each
> constituent that you put onto the parse chart. Then you can check each
> constituent as to whether its presuppositions are fulfilled in the given
> situation.

That method is similar to the way the VivoMind system constructs
the interconnected conceptual graphs that represent a document.
Most of the graphs are derived from the text that is being parsed,
but others come from any source that can contribute anything useful:

    http://www.jfsowa.com/pubs/paradigm.pdf
    Two paradigms are better than one,
    and multiple paradigms are even better

The VivoMind system also uses some ideas that are related to the
methods of the following references cited in that article:

Hahn, Udo, Susanne Schacht, & Norbert Bröker (1994) Concurrent natural
language parsing: The ParseTalk model, International Journal of Human-
Computer Studies, vol 41, pp. 179-222.

Bröker, Norbert (1999) Eine Dependenzgrammatik zur Kopplung heterogener
Wissensquellen, Tübingen: Max Niemeyer Verlag.

The VivoMind system has not yet been applied to robotics, but it has
been adapted to several widely different domains.  For applications,
see Section 7 (Slides 143 to 164) of

    http://www.jfsowa.com/talks/goal.pdf
    The goal of language understanding

Section 6 (Slides 118 to 142) discusses the use of analogies and case-
based reasoning.  In particular, Slides 121 to 126 show how the analogy
engine can relate conceptual graphs derived from different sources that
use different ontologies:  a structure of blocks as represented in a
database and as described in English sentences.

John Sowa

_______________________________________________
UNSUBSCRIBE from this page: http://mailman.uib.no/options/corpora
Corpora mailing list
Corpora at uib.no
http://mailman.uib.no/listinfo/corpora