Answer me this: Constructing Disambiguation Queries for Explanation Generation in Robotics

Tiago Mota; Mohan Sridharan

doi:10.1109/ICDL49984.2021.9515605

Answer me this: Constructing Disambiguation Queries for Explanation Generation in Robotics

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

Our architecture seeks to enable robots collaborating with humans to describe their decisions and evolution of beliefs. To achieve the desired transparency in integrated robot systems that support knowledge-based reasoning and data-driven learning, we build on a baseline system that supports non-monotonic logical reasoning with incomplete commonsense domain knowledge, data-driven learning from a limited set of examples, and inductive learning of previously unknown axioms governing domain dynamics. In the context of a simulated robot providing on-demand, relational descriptions as explanations of its decisions and beliefs, we introduce an interactive system that automatically traces beliefs, and addresses ambiguity in the human queries by constructing and posing suitable disambiguation queries. We present results of evaluation in scene understanding and planning tasks to demonstrate our architecture's abilities.

Original language	English
Title of host publication	IEEE International Conference on Development and Learning, ICDL 2021
Publisher	Institute of Electrical and Electronics Engineers
Number of pages	8
ISBN (Electronic)	9781728162423
DOIs	https://doi.org/10.1109/ICDL49984.2021.9515605
Publication status	Published - 23 Aug 2021
Event	2021 IEEE International Conference on Development and Learning, ICDL 2021 - Virtual, Beijing, China Duration: 23 Aug 2021 → 26 Aug 2021

Conference

Conference	2021 IEEE International Conference on Development and Learning, ICDL 2021
Country/Territory	China
City	Virtual, Beijing
Period	23/08/21 → 26/08/21

Keywords / Materials (for Non-textual outputs)

Deep learning
Explainable reasoning and learning
HRI
Non-monotonic logical reasoning

Access to Document

10.1109/ICDL49984.2021.9515605Licence: All Rights Reserved

https://ieeexplore.ieee.org/document/9515605Licence: All Rights Reserved

Cite this

@inproceedings{7af9e61c95af454e8d311f2992bab9db,

title = "Answer me this: Constructing Disambiguation Queries for Explanation Generation in Robotics",

abstract = "Our architecture seeks to enable robots collaborating with humans to describe their decisions and evolution of beliefs. To achieve the desired transparency in integrated robot systems that support knowledge-based reasoning and data-driven learning, we build on a baseline system that supports non-monotonic logical reasoning with incomplete commonsense domain knowledge, data-driven learning from a limited set of examples, and inductive learning of previously unknown axioms governing domain dynamics. In the context of a simulated robot providing on-demand, relational descriptions as explanations of its decisions and beliefs, we introduce an interactive system that automatically traces beliefs, and addresses ambiguity in the human queries by constructing and posing suitable disambiguation queries. We present results of evaluation in scene understanding and planning tasks to demonstrate our architecture's abilities.",

keywords = "Deep learning, Explainable reasoning and learning, HRI, Non-monotonic logical reasoning",

author = "Tiago Mota and Mohan Sridharan",

note = "Publisher Copyright: {\textcopyright} 2021 IEEE.; 2021 IEEE International Conference on Development and Learning, ICDL 2021 ; Conference date: 23-08-2021 Through 26-08-2021",

year = "2021",

month = aug,

day = "23",

doi = "10.1109/ICDL49984.2021.9515605",

language = "English",

booktitle = "IEEE International Conference on Development and Learning, ICDL 2021",

publisher = "Institute of Electrical and Electronics Engineers",

address = "United States",

}

Mota, T & Sridharan, M 2021, Answer me this: Constructing Disambiguation Queries for Explanation Generation in Robotics. in IEEE International Conference on Development and Learning, ICDL 2021. Institute of Electrical and Electronics Engineers, 2021 IEEE International Conference on Development and Learning, ICDL 2021, Virtual, Beijing, China, 23/08/21. https://doi.org/10.1109/ICDL49984.2021.9515605

TY - GEN

T1 - Answer me this

T2 - 2021 IEEE International Conference on Development and Learning, ICDL 2021

AU - Mota, Tiago

AU - Sridharan, Mohan

PY - 2021/8/23

Y1 - 2021/8/23

N2 - Our architecture seeks to enable robots collaborating with humans to describe their decisions and evolution of beliefs. To achieve the desired transparency in integrated robot systems that support knowledge-based reasoning and data-driven learning, we build on a baseline system that supports non-monotonic logical reasoning with incomplete commonsense domain knowledge, data-driven learning from a limited set of examples, and inductive learning of previously unknown axioms governing domain dynamics. In the context of a simulated robot providing on-demand, relational descriptions as explanations of its decisions and beliefs, we introduce an interactive system that automatically traces beliefs, and addresses ambiguity in the human queries by constructing and posing suitable disambiguation queries. We present results of evaluation in scene understanding and planning tasks to demonstrate our architecture's abilities.

AB - Our architecture seeks to enable robots collaborating with humans to describe their decisions and evolution of beliefs. To achieve the desired transparency in integrated robot systems that support knowledge-based reasoning and data-driven learning, we build on a baseline system that supports non-monotonic logical reasoning with incomplete commonsense domain knowledge, data-driven learning from a limited set of examples, and inductive learning of previously unknown axioms governing domain dynamics. In the context of a simulated robot providing on-demand, relational descriptions as explanations of its decisions and beliefs, we introduce an interactive system that automatically traces beliefs, and addresses ambiguity in the human queries by constructing and posing suitable disambiguation queries. We present results of evaluation in scene understanding and planning tasks to demonstrate our architecture's abilities.

KW - Deep learning

KW - Explainable reasoning and learning

KW - HRI

KW - Non-monotonic logical reasoning

U2 - 10.1109/ICDL49984.2021.9515605

DO - 10.1109/ICDL49984.2021.9515605

M3 - Conference contribution

AN - SCOPUS:85114554297

BT - IEEE International Conference on Development and Learning, ICDL 2021

PB - Institute of Electrical and Electronics Engineers

Y2 - 23 August 2021 through 26 August 2021

ER -

Answer me this: Constructing Disambiguation Queries for Explanation Generation in Robotics

Abstract

Conference

Keywords / Materials (for Non-textual outputs)

Access to Document

Fingerprint

Cite this