Explainable Autonomy: A Study of Explanation Styles for Building Clear Mental Models

Francisco Javier Chiyah Garcia; David A. Robb; Xingkun Liu; Atanas  Laskov; Pedro Patron; Helen Hastie

Explainable Autonomy: A Study of Explanation Styles for Building Clear Mental Models

Francisco Javier Chiyah Garcia, David A. Robb, Xingkun Liu, Atanas Laskov, Pedro Patron, Helen Hastie

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

22 Citations (Scopus)

314 Downloads (Pure)

Abstract

As unmanned vehicles become more autonomous, it is important to maintain a high level of transparency regarding their behaviour and how they operate. This is particularly important in remote locations where they cannot be directly observed. Here, we describe a method for generating explanations in natural language of autonomous system behaviour and reasoning. Our method involves deriving an interpretable model of autonomy through having an expert ‘speak aloud’ and providing various levels of detail based on this model. Through an online evaluation study with operators, we show it is best to generate explanations with multiple possible reasons but tersely worded. This work has implications for designing interfaces for autonomy as well as for explainable AI and operator training.

Original language	English
Title of host publication	Proceedings of the 11th International Conference on Natural Language Generation
Publisher	Association for Computational Linguistics
Pages	99-108
Number of pages	10
ISBN (Electronic)	9781948087865
Publication status	Published - Nov 2018
Event	11th International Conference of Natural Language Generation 2018 - Tilburg University, Tilburg, Netherlands Duration: 5 Nov 2016 → 8 Nov 2018 https://inlg2018.uvt.nl/

Conference

Conference	11th International Conference of Natural Language Generation 2018
Abbreviated title	INLG'18
Country/Territory	Netherlands
City	Tilburg
Period	5/11/16 → 8/11/18
Internet address	https://inlg2018.uvt.nl/

Access to Document

Download pdf
(c) 2018 Association for Computational Linguistics
Accepted author manuscript, 1.48 MB

http://aclweb.org/anthology/W18-6511

Cite this

@inproceedings{724391c106d3475ea4bb8a594a90d031,

title = "Explainable Autonomy: A Study of Explanation Styles for Building Clear Mental Models",

abstract = "As unmanned vehicles become more autonomous, it is important to maintain a high level of transparency regarding their behaviour and how they operate. This is particularly important in remote locations where they cannot be directly observed. Here, we describe a method for generating explanations in natural language of autonomous system behaviour and reasoning. Our method involves deriving an interpretable model of autonomy through having an expert {\textquoteleft}speak aloud{\textquoteright} and providing various levels of detail based on this model. Through an online evaluation study with operators, we show it is best to generate explanations with multiple possible reasons but tersely worded. This work has implications for designing interfaces for autonomy as well as for explainable AI and operator training.",

author = "Garcia, {Francisco Javier Chiyah} and Robb, {David A.} and Xingkun Liu and Atanas Laskov and Pedro Patron and Helen Hastie",

year = "2018",

month = nov,

language = "English",

pages = "99--108",

booktitle = "Proceedings of the 11th International Conference on Natural Language Generation",

publisher = "Association for Computational Linguistics",

note = "11th International Conference of Natural Language Generation 2018, INLG'18 ; Conference date: 05-11-2016 Through 08-11-2018",

url = "https://inlg2018.uvt.nl/",

}

Garcia, FJC, Robb, DA, Liu, X, Laskov, A, Patron, P & Hastie, H 2018, Explainable Autonomy: A Study of Explanation Styles for Building Clear Mental Models. in Proceedings of the 11th International Conference on Natural Language Generation. Association for Computational Linguistics, pp. 99-108, 11th International Conference of Natural Language Generation 2018, Tilburg, Netherlands, 5/11/16. <http://aclweb.org/anthology/W18-6511>

Explainable Autonomy: A Study of Explanation Styles for Building Clear Mental Models. / Garcia, Francisco Javier Chiyah; Robb, David A.; Liu, Xingkun et al.
Proceedings of the 11th International Conference on Natural Language Generation. Association for Computational Linguistics, 2018. p. 99-108.

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

TY - GEN

T1 - Explainable Autonomy: A Study of Explanation Styles for Building Clear Mental Models

AU - Garcia, Francisco Javier Chiyah

AU - Robb, David A.

AU - Liu, Xingkun

AU - Laskov, Atanas

AU - Patron, Pedro

AU - Hastie, Helen

PY - 2018/11

Y1 - 2018/11

N2 - As unmanned vehicles become more autonomous, it is important to maintain a high level of transparency regarding their behaviour and how they operate. This is particularly important in remote locations where they cannot be directly observed. Here, we describe a method for generating explanations in natural language of autonomous system behaviour and reasoning. Our method involves deriving an interpretable model of autonomy through having an expert ‘speak aloud’ and providing various levels of detail based on this model. Through an online evaluation study with operators, we show it is best to generate explanations with multiple possible reasons but tersely worded. This work has implications for designing interfaces for autonomy as well as for explainable AI and operator training.

AB - As unmanned vehicles become more autonomous, it is important to maintain a high level of transparency regarding their behaviour and how they operate. This is particularly important in remote locations where they cannot be directly observed. Here, we describe a method for generating explanations in natural language of autonomous system behaviour and reasoning. Our method involves deriving an interpretable model of autonomy through having an expert ‘speak aloud’ and providing various levels of detail based on this model. Through an online evaluation study with operators, we show it is best to generate explanations with multiple possible reasons but tersely worded. This work has implications for designing interfaces for autonomy as well as for explainable AI and operator training.

M3 - Conference contribution

SP - 99

EP - 108

BT - Proceedings of the 11th International Conference on Natural Language Generation

PB - Association for Computational Linguistics

T2 - 11th International Conference of Natural Language Generation 2018

Y2 - 5 November 2016 through 8 November 2018

ER -

Explainable Autonomy: A Study of Explanation Styles for Building Clear Mental Models

Abstract

Conference

Access to Document

Fingerprint

Cite this