Information Technology Reference
In-Depth Information
Ta b l e 1 .
Survey results by sentence type
absolute
relative
type
frequency frequency
imperatives
114
87%
imbedded imperatives
6
5%
need-statements
2
2%
4
3%
hints
wh-questions
3
2%
3
2%
others
4.1
Understanding
The aim of our approach was to provide a system that is able to react to as many com-
mands for a domestic service robot given in natural language as possible. With the
generic grammar for English directives our approach is able to handle more utterances
than previous approaches based on finite state grammars such as [7]. To evaluate how
far off we are from an ideal natural language interface we conducted a user survey.
The survey was carried out on-line with a small group of (about 15) predominantly
tech-savvy students. A short description of the robot's capabilities was given and par-
ticipants were asked to provide us with sample requests for our system. Participants
took the survey without any assistance, except the task description.
We received a total of 132 submissions. Firstly, we are interested in the general struc-
ture of the answers to see whether our grammar is appropriate. Therefore, Table 1 shows
the submissions itemised by sentence type.
Syntactically speaking, the grammar can cover imperatives, imbedded imperatives
and need-statements, which make for 92.37% of the survey results. However, some of
these utterances do not possess the verb-object-structure we assumed in our system.
For example, “Make me a coffee the way I like it” contained an adverbial (“the way I
like it”) which we did not account for neither in the grammar nor in the interpretation
process. It is technically possible to treat adverbials as entities and thus incorporate such
utterances. A better founded approach, however, would be to introduce the concept of
adverbials to our system as a special case of objects that modify the mode of a skill.
We leave this for future work, though. Still, 77.01% of the survey entries provide the
assumed modular verb-object-structure and can therefore be processed by our system
successfully.
To test the resilience against erroneous utterances we tested the system's response to
the set of utterances given in Table 2. In case that an object is missing that is required
as a parameter by a skill (as in E1) the system will inquire for clarification by offer-
ing possible entities. To be able to handle unspecific objects we included those in our
grammar and we treat them just like missing objects and initiate a clarification proce-
dure. Preposition help in assigning objects to parameter slots of a skill. With only one
parameter as in the utterance E3 we do not require the preposition in order to come
to a successful termination of our interpretation process. With multiple parameters that
are identical in all their attributes we would need additional information though. We
do not make use of prepositions to resolve these kinds of confusions, yet. An utterance
