Are contrastive explanations useful?

James Forrest; Somayajulu Sripada; Wei Pang; George M. Coghill

Are contrastive explanations useful?

James Forrest, Somayajulu Sripada, Wei Pang, George M. Coghill

Heriot-Watt University

Research output: Contribution to conference › Unpublished paper › peer-review

1 Citation (Scopus)

38 Downloads (Pure)

Abstract

From the user perspective (data subjects and data controllers), useful explanations of ML decisions are selective, contrastive and social. In this paper, we describe an algorithm for generating selective and contrastive explanations and experimentally study its usefulness to users.

Original language	English
Pages	9-16
Number of pages	8
Publication status	Published - 1 Jun 2021
Event	2021 SICSA eXplainable Artifical Intelligence Workshop, SICSA XAI 2021 - Aberdeen, United Kingdom Duration: 1 Jun 2021 → 1 Jun 2021

Conference

Conference	2021 SICSA eXplainable Artifical Intelligence Workshop, SICSA XAI 2021
Country/Territory	United Kingdom
City	Aberdeen
Period	1/06/21 → 1/06/21

Bibliographical note

Funding Information:
Supported by EPSRC DTP Grant Number EP/N509814/1

Keywords

Contrastive explanations
Interpretable ML
XAI

Access to Document

Forrest_etal_are_contrastive_explanations_VOR
Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0). https://creativecommons.org/licenses/by/4.0/
Final published version, 554 KBLicence: CC BY

Cite this

@conference{b88f5792cf0c4a0a94ea1dfe22f76b09,

title = "Are contrastive explanations useful?",

abstract = "From the user perspective (data subjects and data controllers), useful explanations of ML decisions are selective, contrastive and social. In this paper, we describe an algorithm for generating selective and contrastive explanations and experimentally study its usefulness to users.",

keywords = "Contrastive explanations, Interpretable ML, XAI",

author = "James Forrest and Somayajulu Sripada and Wei Pang and Coghill, {George M.}",

note = "Funding Information: Supported by EPSRC DTP Grant Number EP/N509814/1 ; 2021 SICSA eXplainable Artifical Intelligence Workshop, SICSA XAI 2021 ; Conference date: 01-06-2021 Through 01-06-2021",

year = "2021",

month = jun,

day = "1",