Information Technology Reference
In-Depth Information
s
−−>
salutation
utterance
|
utterance
%
utterance
−−>
needstatement
|
imperative
|
imbedded imperative
%
needstatement
−−>
np
vp
|
needphrase
vp
imperative
−−>
vp
imbedded imperative
−−>
aux np vp
needphrase
−−>
”i” prompt ”you to”
% verb
phrase
vp
−−>
vp '
|
vp'
conjunction
vp
vp '
−−>
verb
|
verb
obp
|
courtesy
vp '
% object
phrase
obp
−−>
np
|
pp
|
np
obp
|
pp
obp
% noun
phrase
np
−−>
noun
|
pronoun
|
determiner
noun
% propositional
phrase
pp
−−>
prep
np
Fig. 2.
Base grammar in EBNF
politeness, the user can express courtesy by saying “please”. Putting all this together,
we arrive at a base grammar that can be expressed in Extended Backus-Naur Form
(EBNF)[18]asshowninFigure2.
In addition to the base grammar we need a base lexicon that provides us with the
vocabulary for elements such as prepositions, auxiliary verbs, courtesies, conjunctions,
determiners, and pronouns. To generate a system that is functional in a specific setting,
we further need a lexicon containing all verbs for the capabilities of the robot as well as
all the objects referring to known entities in the world. This depends on the particular
application, though. That is why we couple this to the domain specification discussed
later. The base grammar, the base lexicon, and the domain specific lexicon then yield
the final grammar that is used for syntactical processing.
Since we are only interested in the core information, the most relevant parts of the
utterance are verbs, objects, prepositions, and determiners. We can drop auxiliary verbs,
filler words, courtesies, and alike without losing any relevant information. Doing so, we
finally arrive at an internal representation of the utterance in a prefix notation depicted
below, that we use for further processing.
[
and
, [[Verb, [
objects
, [[Preposition, [Determiner,Object]],...]] ]], ...]
The list notation contains the keyword
and
to concatenate multiple verb phrases and
it uses the keyword
objects
to group the object phrase. If an utterance is missing,
information we fill this with
nil
as a placeholder.
3.2
Planning Interpretations
After syntactic pre-processing of an utterance into the internal representation, the sys-
tem uses decision-theoretic planning to arrive at the most likely interpretation of the
utterance, given the robot's capabilities. The interpretation is supposed to match the re-
quest with one of the abilities of the robot (called a skill) and to correctly allocate the
parameters that this skill requires.
In order to do that, we need to identify the skill that is being addressed first. We are
going about this from the verb which has been extracted in the syntactical processing,
