Skip to main navigation Skip to search Skip to main content

Interpreting route instructions as qualitative spatial actions

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

In this paper we motivate the use of qualitative spatial actions as the fundamental unit in processing user route instructions. The spatial action model has been motivated by an analysis of empirical studies in human-robot interaction on the navigation task, and can be interpreted as a conceptual representation of the spatial action to be performed by the agent in their navigation space. Furthermore, we sketch out two distinct models of interpretation for these actions in cognitive robotics. In the first, the actions are related to a formalized conceptual user modeling of navigation space, while in the second the actions are interpreted as fuzzy operations on a voronoi graph. Moreover, we show how this action model allows us to better capture the points at which user route instructions become out of alignment with a robot's knowledge of the environment through a number of examples.

Original languageEnglish
Title of host publicationSpatial Cognition V
Subtitle of host publicationReasoning, Action, Interaction - International Conference Spatial Cognition 2006, Revised Selected Papers
PublisherSpringer Verlag
Pages327-345
Number of pages19
ISBN (Print)9783540756651
DOIs
Publication statusPublished - 2007
Externally publishedYes
Event5th International Conference Spatial Cognition 2006 - Bremen, Germany
Duration: 24 Sep 200628 Sep 2006

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume4387 LNAI
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference5th International Conference Spatial Cognition 2006
Country/TerritoryGermany
CityBremen
Period24/09/0628/09/06

Fingerprint

Dive into the research topics of 'Interpreting route instructions as qualitative spatial actions'. Together they form a unique fingerprint.

Cite this