Using the toulmin model of argumentation to explore the differences in human and automated hiring decisions

Hebah Bubakr; Chris Baber

Using the toulmin model of argumentation to explore the differences in human and automated hiring decisions

Engineering and Physical Sciences

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

Abstract

Amazon developed an experimental hiring tool, using AI to review job applicants' résumés, with the goal of automating the search for the best talent. However, the team found that their software was biased against women because the models were trained on résumés submitted to the company for the previous 10 years and most of these were submitted by men, reflecting male dominance in the tech business. As a result, the models learned that males were preferable, and it excluded résumés that could be inferred to come from female applicants. Gender bias was not the only issue. As well rejecting plausible candidates, problems with the data lead the models to recommend unqualified candidates for jobs. To understand the conflict in this, and similar examples, we apply Toulmin model of argumentation. By considering how arguments are constructed by a human and how a contrasting argument might be constructed by AI, we can conduct pre-mortems of potential conflict in system operation.

Original language	English
Title of host publication	HUCAPP
Editors	Manuela Chessa, Alexis Paljic, Jose Braz
Publisher	SciTePress
Pages	211-216
Number of pages	6
ISBN (Electronic)	9789897584022
Publication status	Published - 2020
Event	15th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications, VISIGRAPP 2020 - Valletta, Malta Duration: 27 Feb 2020 → 29 Feb 2020

Publication series

Name	VISIGRAPP 2020 - Proceedings of the 15th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications
Volume	2

Conference

Conference	15th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications, VISIGRAPP 2020
Country/Territory	Malta
City	Valletta
Period	27/02/20 → 29/02/20

Bibliographical note

Publisher Copyright:
Copyright © 2020 by SCITEPRESS - Science and Technology Publications, Lda. All rights reserved.

Keywords

Artificial intelligence
Human computer interaction
Toulmin model of argumentation

ASJC Scopus subject areas

Computer Graphics and Computer-Aided Design
Computer Science Applications
Computer Vision and Pattern Recognition

Cite this

Bubakr, H., & Baber, C. (2020). Using the toulmin model of argumentation to explore the differences in human and automated hiring decisions. In M. Chessa, A. Paljic, & J. Braz (Eds.), HUCAPP (pp. 211-216). (VISIGRAPP 2020 - Proceedings of the 15th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications; Vol. 2). SciTePress.

Bubakr, Hebah ; Baber, Chris. / Using the toulmin model of argumentation to explore the differences in human and automated hiring decisions. HUCAPP. editor / Manuela Chessa ; Alexis Paljic ; Jose Braz. SciTePress, 2020. pp. 211-216 (VISIGRAPP 2020 - Proceedings of the 15th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications).

@inproceedings{696c4db4aaad46208d9f8f5f42b488f2,

title = "Using the toulmin model of argumentation to explore the differences in human and automated hiring decisions",

abstract = "Amazon developed an experimental hiring tool, using AI to review job applicants' r{\'e}sum{\'e}s, with the goal of automating the search for the best talent. However, the team found that their software was biased against women because the models were trained on r{\'e}sum{\'e}s submitted to the company for the previous 10 years and most of these were submitted by men, reflecting male dominance in the tech business. As a result, the models learned that males were preferable, and it excluded r{\'e}sum{\'e}s that could be inferred to come from female applicants. Gender bias was not the only issue. As well rejecting plausible candidates, problems with the data lead the models to recommend unqualified candidates for jobs. To understand the conflict in this, and similar examples, we apply Toulmin model of argumentation. By considering how arguments are constructed by a human and how a contrasting argument might be constructed by AI, we can conduct pre-mortems of potential conflict in system operation.",

keywords = "Artificial intelligence, Human computer interaction, Toulmin model of argumentation",

author = "Hebah Bubakr and Chris Baber",

note = "Publisher Copyright: Copyright {\textcopyright} 2020 by SCITEPRESS - Science and Technology Publications, Lda. All rights reserved.; 15th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications, VISIGRAPP 2020 ; Conference date: 27-02-2020 Through 29-02-2020",

year = "2020",

language = "English",

series = "VISIGRAPP 2020 - Proceedings of the 15th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications",

publisher = "SciTePress",

pages = "211--216",

editor = "Manuela Chessa and Alexis Paljic and Jose Braz",

booktitle = "HUCAPP",

}

Bubakr, H & Baber, C 2020, Using the toulmin model of argumentation to explore the differences in human and automated hiring decisions. in M Chessa, A Paljic & J Braz (eds), HUCAPP. VISIGRAPP 2020 - Proceedings of the 15th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications, vol. 2, SciTePress, pp. 211-216, 15th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications, VISIGRAPP 2020, Valletta, Malta, 27/02/20.

Using the toulmin model of argumentation to explore the differences in human and automated hiring decisions. / Bubakr, Hebah; Baber, Chris.
HUCAPP. ed. / Manuela Chessa; Alexis Paljic; Jose Braz. SciTePress, 2020. p. 211-216 (VISIGRAPP 2020 - Proceedings of the 15th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications; Vol. 2).

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

TY - GEN

T1 - Using the toulmin model of argumentation to explore the differences in human and automated hiring decisions

AU - Bubakr, Hebah

AU - Baber, Chris

PY - 2020

Y1 - 2020

N2 - Amazon developed an experimental hiring tool, using AI to review job applicants' résumés, with the goal of automating the search for the best talent. However, the team found that their software was biased against women because the models were trained on résumés submitted to the company for the previous 10 years and most of these were submitted by men, reflecting male dominance in the tech business. As a result, the models learned that males were preferable, and it excluded résumés that could be inferred to come from female applicants. Gender bias was not the only issue. As well rejecting plausible candidates, problems with the data lead the models to recommend unqualified candidates for jobs. To understand the conflict in this, and similar examples, we apply Toulmin model of argumentation. By considering how arguments are constructed by a human and how a contrasting argument might be constructed by AI, we can conduct pre-mortems of potential conflict in system operation.

AB - Amazon developed an experimental hiring tool, using AI to review job applicants' résumés, with the goal of automating the search for the best talent. However, the team found that their software was biased against women because the models were trained on résumés submitted to the company for the previous 10 years and most of these were submitted by men, reflecting male dominance in the tech business. As a result, the models learned that males were preferable, and it excluded résumés that could be inferred to come from female applicants. Gender bias was not the only issue. As well rejecting plausible candidates, problems with the data lead the models to recommend unqualified candidates for jobs. To understand the conflict in this, and similar examples, we apply Toulmin model of argumentation. By considering how arguments are constructed by a human and how a contrasting argument might be constructed by AI, we can conduct pre-mortems of potential conflict in system operation.

KW - Artificial intelligence

KW - Human computer interaction

KW - Toulmin model of argumentation

UR - http://www.scopus.com/inward/record.url?scp=85083507401&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:85083507401

T3 - VISIGRAPP 2020 - Proceedings of the 15th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications

SP - 211

EP - 216

BT - HUCAPP

A2 - Chessa, Manuela

A2 - Paljic, Alexis

A2 - Braz, Jose

PB - SciTePress

T2 - 15th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications, VISIGRAPP 2020

Y2 - 27 February 2020 through 29 February 2020

ER -

Using the toulmin model of argumentation to explore the differences in human and automated hiring decisions

Abstract

Publication series

Conference

Bibliographical note

Keywords

ASJC Scopus subject areas

Fingerprint

Cite this