Symbol Grounding and Task Learning from Imperfect Corrections

Mattias Appelgren, Alex Lascarides


Abstract
This paper describes a method for learning from a teacher’s potentially unreliable corrective feedback in an interactive task learning setting. The graphical model uses discourse coherence to jointly learn symbol grounding, domain concepts and valid plans. Our experiments show that the agent learns its domain-level task in spite of the teacher’s mistakes.
Anthology ID:
2021.splurobonlp-1.1
Volume:
Proceedings of Second International Combined Workshop on Spatial Language Understanding and Grounded Communication for Robotics
Month:
August
Year:
2021
Address:
Online
Editors:
Malihe Alikhani, Valts Blukis, Parisa Kordjamshidi, Aishwarya Padmakumar, Hao Tan
Venue:
splurobonlp
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
1–10
Language:
URL:
https://aclanthology.org/2021.splurobonlp-1.1
DOI:
10.18653/v1/2021.splurobonlp-1.1
Bibkey:
Cite (ACL):
Mattias Appelgren and Alex Lascarides. 2021. Symbol Grounding and Task Learning from Imperfect Corrections. In Proceedings of Second International Combined Workshop on Spatial Language Understanding and Grounded Communication for Robotics, pages 1–10, Online. Association for Computational Linguistics.
Cite (Informal):
Symbol Grounding and Task Learning from Imperfect Corrections (Appelgren & Lascarides, splurobonlp 2021)
Copy Citation:
PDF:
https://aclanthology.org/2021.splurobonlp-1.1.pdf