Communicating with Executable Action Representations

TitleCommunicating with Executable Action Representations
Publication TypeConference Paper
Year of Publication2013
AuthorsSchilling, M., & Narayanan S.
Other Numbers3406
Abstract

Natural language instructions are often underspecified andimprecise which makes them hard to understand for an artificialagent. In this article we present a system of connectedknowledge representations that is used to control a robotthrough instructions. As actions are a key component of instructionsand the robot’s behavior the representation of actionis central in our approach. First, the system consists ofa conceptual schema representation which provides a parameterinterface for action. Second, we present an intermediaterepresentation of the temporal structure of action and showhow this generic action structure can be mapped to detailedaction controllers as well as language.

Acknowledgment

This work was partially funded by the Center of Excellence Cognitive Interaction Technology (EXC 277) and by the Deutscher Akademischer Austausch Dienst (DAAD) through a postdoctoral fellowship. Additional funding was provided by the Office of Naval Research (ONR). Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors or originators and do not necessarily reflect the views of ONR or of the U.S. Government.

URLhttp://www.icsi.berkeley.edu/pubs/ai/executableactionreps13.pdf
Bibliographic Notes

Proceedings of the AAAI Spring Symposium (Designing Intelligent Robots), Stanford University, California

Abbreviated Authors

M. Schilling and S. Narayanan

ICSI Research Group

AI

ICSI Publication Type

Article in conference proceedings